AI Papers · Daily Picks

LingBot-Touch is a tactile foundation model that learns a unified, extensible representation across heterogeneous tactile sensors, yielding plug-and-play tactile tokens that consistently enhance manipulation policies such as diffusion policy and VLA models.

2611.12806 · ▲ 102 · Code

A full-context speech recognizer is adapted to emit immutable streaming transcripts with near full-context accuracy and low latency by learning longest stable prefix labels.

2611.06043 · ▲ 92 · Code

APEX extends a streaming video generation method by replacing a separate fake-score network with condition shifting to estimate the generated distribution’s velocity field, improving instruction following and visual aesthetics while reducing memory and iteration time relative to DMD.

2611.14033 · ▲ 88

SplatAnything applies Gaussian Spacing Normalization to canonicalize scene-level Gaussian spacing, yielding more consistent supervision and improved feed-forward 3D Gaussian Splatting under varying scene scales and viewpoints.

2611.08264 · ▲ 69 · Code

P^2-MOPD jointly optimizes mixing coefficients and student weights under an on-policy distillation objective, enabling faster and more balanced recovery than MOPD by treating experts as parameter priors while keeping the student trainable.

2611.03953 · ▲ 69 · Code

WikiTrace links each distilled lesson to its source trajectory and rationale, enabling LLM agents to verify and correct memories before acting, improving performance and reducing memory-induced degradation across agentic benchmarks.

2611.14320 · ▲ 62 · Code

Tree-Ising repairs reasoning loops in LLM weights by solving sparse Ising/QUBO problems over BF16 parameters, improving math benchmark accuracy with fewer weight edits and no gradient or LLM-token cost.

2611.14210 · ▲ 62 · Code

WorldFoundry offers open-source infrastructure for reproducible inference and evaluation of generative world models across video, 3D/4D, embodied action, and interactive systems, with traceable evidence and benchmarks.

2611.10771 · ▲ 58 · Code

Trained action policies write manipulation skill into network weights, tying what is learned to the distribution it is learned from. In-context learning (ICL) avoids this coupling but has not matched trained policies, because a demonstration recorded in another scene carries no feedback from the world the robot currently faces. We observe that what an agent learns in context is semantic, recording task steps and object roles rather than visual appearance, so experience gathered in one world transfers to another. Based on this insight, we propose World In-Context Learning (WICL), a paradigm in which a frozen model starts from a demonstration, rehearses the task in a world such as a simulator, and writes the feedback into its context rather than its weights. We build RoboScript, a modular infrastructure that implements WICL for agentic manipulation. From a single demonstration, RoboScript learns each task through rehearsal in simulation, freezes the learned knowledge as a Task Memory, and deploys the frozen memory on new scenes with data isolation that ensures every deployment input is available on a real robot. On all 50 tasks of RoboTwin 2.0, RoboScript reaches 83.4% mean success across clean and randomized settings with one demonstration per task and no gradient updates, exceeding the strongest trained action policy by over 20 points. The method generalizes across visual distributions, improves through reflection, scales with the foundation model, and transfers to a physical dual-arm robot through cross-embodiment memory adaptation. With a world in the loop, a robot can learn a manipulation task in its context as well as it learns one in its weights. All code is available at https://github.com/ArcheOne-AI/RoboScript

2611.12885 · ▲ 58 · Code

We introduce Sage-Taiyi, a metacognition-inspired architecture that uses accumulated scientific understanding to guide what to investigate next and how to investigate it. Through explicit monitoring and control, the system revises its understanding of evidence, assumptions, and unresolved questions, then uses these assessments to adapt research targets and select reusable reasoning methods. On Taiyi-Bench, an AI co-scientist ideation benchmark comprising 265 tasks from 50 papers across ten disciplines, Sage-Taiyi ranks first on all eight metrics of historical rediscovery and open-ended ideation, achieving an Overall score of 54.0 compared with 47.9 for the strongest baseline, AI Scientist-v2. It also performs well on two external scientific ideation benchmarks. Five expert-guided collaborations further connect its proposals to experimental and computational validation across chemistry, materials science, and engineering. In chemistry, reaction-condition optimization raises the yield of cobalt-catalyzed carbonylation from trace amounts to 84% and photocatalytic allylic amination from 13% to 69%. These results support metacognitive regulation as an approach to scientific ideation and its progression toward practical validation.

2611.12762 · ▲ 56 · Code

An embodied multi-agent framework uses learned correction codes and a vision-language-action model with motion-aware history and latent world modeling to improve robot control, failure recovery, and training efficiency.

2611.09124 · ▲ 53 · Code

Automating printed circuit board (PCB) routing requires coordinating board-level connectivity decisions with geometric execution under strict design rules. Although recent agents can interact with native routing tools, reliable routing on complex boards remains challenging. We present PCBRouteAgent, a multimodal PCB routing agent inspired by how human engineers coordinate global strategy with local execution. A Global Routing Planner combines board images and structured electrical and geometric information to specify connection topology, routing order, corridors, and layer transitions. A Detailed Routing Executor realizes these guides through KiCad's native routing engine and performs bounded local recovery. Unresolved failures are returned as structured feedback, enabling targeted guide revisions while preserving accepted routes. We evaluate PCBRouteAgent on 247 synthetic and real-world boards from four PCBWorld-Bench test splits. It achieves clean-pass rates of 100%, 94.9%, 100%, and 40% on D2, D3-A, D3-B, and D3-C, respectively. On the three real-world subsets, PCBRouteAgent outperforms the best evaluated baseline in clean-pass rate by 13.9, 36, and 10 percentage points, respectively.

2611.08175 · ▲ 42 · Code

A unified decoder-only policy guided by reward-model-based reinforcement learning generates recommendation exposure slates end-to-end, improving business metrics and reducing reranking and mix-ranking latency in a large-scale system.

2611.04720 · ▲ 41

Reframe is a two-stage post-training framework that aligns visual agents’ observation states through evidence-aware reinforcement learning and observation-gated on-policy distillation, improving tool use and outperforming Mix-RL and vanilla MOPD across image and video benchmarks.

2611.12886 · ▲ 40 · Code

A large-scale corpus of structured proof-supervision records derived from university mathematics textbooks improves language model mathematical reasoning by 16.7 percentage points over a base model when used for fine-tuning.

2610.40394 · ▲ 40

A large aligned benchmark reveals that tabular generation, imputation, and prediction share useful conditional information but show distinct fidelity rankings and objective interactions, motivating unified models that improve categorical reconstruction and joint sampling.

2611.13001 · ▲ 39 · Code

In this work, we study whether frontier agent can continually improve real-world AI infrastructure over long horizons. We introduce Infrathon, a benchmark of diverse, long-horizon AI infrastructure optimization tasks across different levels. We evaluate six models across 150 runs with 24-hour budgets, totaling over 10,000 GPU-hours. Our results show that model rankings vary with the evaluation horizon, while 135 of the 150 runs achieve further improvements during the second half of their budgets. These findings indicate that shorter evaluations can overlook both optimization gains and changes in relative model performance. We also find agents can discover solutions that outperform human-designed baselines: Claude Opus 5.5 increases the mean KV-cache compression ratio from 6.82% with the human-designed Quant-VideoGen codec to 19.86%. Analysis of agent trajectories identifies task violations, unreliable testing, and premature completion as barriers to sustained improvement. Together, Our Infrathon framework provide resources for constructing long-horizon tasks, evaluating frontier agents, and discovering good infrastructure solutions.

2611.11995 · ▲ 36 · Code

Flide is an exact inference engine that accelerates decision-model LLM agents by decoupling logical request trees from physical execution and using ForkAttention to preserve exact branch evaluations.

2610.42148 · ▲ 35 · Code

A discrete native-resolution vision transformer with grouped product quantization and a dual-branch objective acts as a versatile visual tokenizer, matching or exceeding task-specific baselines across visual understanding, generation, and editing tasks.

2611.12426 · ▲ 35

A benchmark of dated U.S. financial facts shows that large language models often understate their knowledge and that common safeguards cannot reliably certify that historical backtests are free of information leakage.

2611.06808 · ▲ 34

Low-bit quantized LLM agents can recover much of their task success through closed-loop guidance from a full-precision teacher, followed by unassisted student training on corrected and reference turns.

2611.13410 · ▲ 32 · Code

While Recursive Self-Improvement (RSI) accelerates isolated algorithm and kernel synthesis, distributed LLM training depends on tightly coupled launch settings, operator kernels, and communication schedules, where single edits often hit out-of-memory (OOM) boundaries, compiler conflicts, measurement jitter, or silent numerical drift. We present InfraRSI, the agentic harness that turns open-ended human commands into verified training optimizations via a five-stage workflow: init locks editable scope and a fixed task contract over workload, loss, and timing invariants to prevent reward hacking, doctor profiles baseline phases and calibrates cluster noise, run co-optimizes launch knobs and code patches before re-verifying top candidates against the baseline and logging findings to a cross-hardware experience bank, and report and apply audit source paths, track failed-trial costs, and commit promoted changes. Rather than stepping greedily from a single incumbent, InfraRSI initializes from diverse experience-seeded configurations and fits global execution-time and censored peak-memory models across all trials, including OOMs. Using these models, it prices per-setting memory costs, revises miscalibrated LLM priors from live runs, and pairs memory-hungry optimizations with memory savers and verified code patches. Across four single- and multi-node dense and Mixture-of-Experts (MoE) LLM workloads, InfraRSI achieves a speedup of up to 2.75times and lifts mean Model FLOPs Utilization (MFU) from 17.4% to 30.0%, outperforming existing tuners and coding agents while transferring experience across GPUs. As it improves training infrastructure efficiency, InfraRSI forms a two-level Recursive Self-Improvement loop---revising priors online within runs and distilling experience across clusters---to push foundation model scaling.

2611.11558 · ▲ 32 · Code

Reproducing machine learning experiments requires reconstructing procedures from information scattered across papers, released code, and supporting artifacts. For coding agents, reproduction success can therefore depend on their ability to fill in missing details, rather than on the released artifacts alone. We ask whether preserving recovered procedures and decisions improves subsequent independent reproduction. We define a standardized artifact design comprising a refined codebase, environment setup, experiment launchers, and a reproduction guide, and introduce ReproBridge, an agent workflow for constructing these refined reproducible artifacts. ReproBridge-Bench provides 1,029 tasks from 387 experiments across 36 papers, anchored by five human-verified reference scores per task. Agents attempt tasks in fresh sessions with raw code or refined artifacts; a separate Auditor checks execution evidence. Across four evaluated models, refined artifacts raise mean reproduction success from 76.4% to 92.6% and reduce downstream API spending by 19.7% on average. The cross-model success-rate spread narrows from 21.6 to 3.9 percentage points, with all four models achieving mean success rates above 90%. Removing any component reduces success in the ablation study, with the largest losses from code and environment removal. These results suggest that preserving execution knowledge helps even models with lower raw-code success reproduce experiments more reliably and at lower API cost.

2611.08452 · ▲ 32

LongBanana is an expert-verified benchmark and agent-harness framework for evaluating and improving long-context multi-reference image synthesis through graph-grounded checklists and an RPR workflow.

2611.06856 · ▲ 31 · Code

MACaQ is an MXFP4 W4A4 post-training quantization method that quantizes each linear layer independently using online activation transforms and Hessian-aware refinement, reducing cost while preserving accuracy across language and diffusion models.

2611.10649 · ▲ 30 · Code

Large language models (LLMs) draw on parametric knowledge and retrieval-augmented generation (RAG) to solve knowledge-intensive tasks. Yet problem solving in unfamiliar domains calls for tightly coupled knowledge acquisition and reasoning, beyond factual retrieval and aggregation or learning solely from a preassembled context. We study self-learning -- autonomously retrieving, studying, and applying knowledge from long-form reference materials on the fly. To investigate this capability, we formulate a textbook-based mathematical problem-solving task. LLMs must navigate textbook exposition, learn and apply relevant principles, and construct proofs for exercises, with reference solutions withheld. We introduce a BookNavigator interface to support this human-inspired study process and incentivize self-learning through reinforcement learning (RL) with fine-grained rewards. Experiments show that our method improves performance across in-domain and out-of-domain textbooks and subjects, raising Intern-S2-Preview-35B's overall proof accuracy from 51.5% to 79.5% and outperforming RL-trained textbook-free and fixed-retrieval baselines. These gains extend to complex long-context benchmarks involving different task types, with improvements over the base model of 1.7 and 2.3 percentage points on BrowseComp-Plus and CL-bench, respectively. Together, these findings suggest that RL can foster a domain-agnostic self-learning strategy, highlighting the potential of self-learning from long-form corpora to strengthen broader problem-solving capabilities on complex, long-context tasks.

2610.39671 · ▲ 28

Early successes on agentic robotics point to a potential paradigm shift toward using coding agents to control robots. Existing robotics benchmarks are built around visuomotor policy learning, leaving it open whether coding agents can reliably control diverse robots under physical-world constraints. We therefore introduce the Embodied First Exam, a unified, comprehensive, and agent-native evaluation framework where coding agents can choose their own control approach: direct control, code to control, or train a policy to control. We hope it serves as a first exam for coding agents before they enter the physical world, with 1,156 tasks in 35 suites on 47 robots, including 9 new suites that go beyond pick-and-place to spatial puzzles, dexterous manipulation, multi-robot interaction, dynamic whole-body control, and sim-to-sim engineering. Each task can be run in up to three evaluation settings. The Standard Setting mimics the real world: the agent receives only sensor observations and, by default, cannot reset the environment. Its Realtime variant keeps the environment running while the agent reasons. The Privileged Setting gives the agent access to the full simulator state and unlimited resets, allowing it to search for an oracle solution for training purposes. Success rates in this setting serve as an approximate upper bound. On the 456 core tasks, the strongest of eight agents, GPT-6 Astra in Codex, achieves a success rate of 34% in the Standard Setting, leaving much room for improvement. To see where agents fail, we label the stage at which failed attempts go wrong and compare failures across agents, settings, and stage categories.

2611.12177 · ▲ 27 · Code

System-1 Decision Models directly make structured probabilistic decisions over query-specified options given a state in a single forward pass, without autoregressively generating reasoning trajectories or textual responses. Recent evaluations cover diverse tasks but often rely on existing benchmarks, where prior exposure to test material can obscure generalization. Moreover, real-world decision making often involves long contexts, large option spaces, and complex logical dependencies, challenges that existing benchmarks do not adequately capture together. Thus, we introduce MMDM (Massive Multitask Decision Making), a benchmark of 5,448 high-quality questions across 16 task categories, combining newly constructed tasks with source-grounded recompositions and rewrites for single-select, multi-select, and soft-label decision making in fresh contexts. It covers realistic scenarios such as agent state tracking, tool and skill selection, preference judgments, and rubric scoring, with option spaces ranging from 2 to 1,024 and inputs extending to 64,817 tokens. With MMDM, we conduct a comprehensive quantitative evaluation of 28 decision models, showing that many perform well on common knowledge yet struggle with interdependent options, constrained decisions, and multi-select tasks. For example, Perplexity Decider-V1.2 reaches 78.75% Hard Single-select Accuracy on world knowledge, but only 53.95% on option logic and 47.95% on constraint planning, while the best Hard Multi-select Accuracy across all models is 36.81%. These results highlight joint reasoning, constraint-aware selection, and complete multi-select recovery as key priorities for decision model development, with MMDM providing a testbed for measuring this progress.

2611.14309 · ▲ 27 · Code

EvoWorld combines a vision-language model that maintains and reasons over world state with a segment-autoregressive video renderer to improve event-conditioned causal evolution and revisit consistency in generative video world simulators.

2611.06748 · ▲ 25 · Code

Video-Coder is a code-driven video generation agent that writes and executes code to build structured 3D draft videos, then uses supervised fine-tuning and reinforcement learning with execution and video rewards to improve controllable video generation.

2611.09894 · ▲ 22 · Code

World action models jointly predict video and actions conditioned on language. Existing studies focus primarily on architectures and training mechanisms, typically treating language-video alignment acquired through video pretraining as an established foundation. However, introducing action learning may impair this alignment. Principled guidance on data and supervision design for maintaining language-video alignment during action learning remains limited. To investigate this, we curate a pool of approximately 47,000 hours of general videos, egocentric videos, UMI demonstrations, and robot demonstrations. Building on this, we systematically study how language-video supervision, language-video-action supervision, and the joint use of these two forms affect WAM performance. In particular, we examine how language annotation quality affects pretrained language-video capabilities, and how action supervision from UMI and robot demonstrations extends these capabilities to language-video-action modeling. To better scale the model, we further conduct detailed studies on joint training recipe and model architecture. Under a fixed training budget, mixing egocentric videos and UMI demonstrations improves control generalization. Model architecture affects how effectively these experiences are used, while scaling pretraining under the selected configuration brings further gains. The resulting recipe, post-trained on RoboTwin Clean demonstrations only, achieves success rates of 88.10% and 61.56% under Easy (Clean-to-Clean) and Hard (Clean-to-Random) conditions, respectively. Our model achieves 40.00% SR across three real-world tasks, compared with 23.33% for the strongest baseline, indicating better generalization under unseen conditions. We release the curated data, training code, and model weights.

2611.12870 · ▲ 22 · Code

Challenging repository-level benchmarks remain limited for many programming language ecosystems, where filtering scarce candidates with ambiguous descriptions or defective tests can discard valuable engineering problems. We introduce Longtail-SWE, a benchmark of 202 issue-resolution tasks across 13 languages with limited coverage in existing repository-level benchmarks: Clojure, Dart, Elixir, Erlang, F#, Haskell, Julia, Kotlin, Lua, OCaml, R, Scala, and Swift. Its construction uses a solver--judge--repairer pipeline: a solver attempts each candidate task, and a judge reviews the task together with the solver's trajectory and test results to guide revisions to tests and task descriptions. These revisions aim to correct tests that reject valid alternative implementations or accept incorrect solutions while retaining challenging tasks and preserving their intended requirements. We evaluate nine model configurations, including frontier models, and analyze task-solving performance, consistency across repeated attempts, and resource usage. We release all benchmark instances, executable environments, complete construction and evaluation tooling, and agent trajectories, together with an additional set of easier tasks to support evaluation of smaller models.

2611.10304 · ▲ 22 · Code

Weight-Space Programming can produce large localized accuracy gains and uneven cross-task transfer, but reliable editing remains inconsistent.

2611.09382 · ▲ 19 · Code

WaveGen is a fully end-to-end diffusion framework for text-to-waveform generation that jointly learns spectral and waveform trajectories via Spectral Forcing without external audio codecs, vocoders, latent models, alignment modules, or self-supervised representations, achieving state-of-the-art text-to-speech results.

2611.12527 · ▲ 16 · Code

How little of a pretrained large language model (LLM) has to change for it to acquire new capabilities? We find that zeroing fewer than 0.05% of its weights, with no other modification, matches full fine-tuning (FFT) at scale across three LLM post-training regimes: reinforcement learning with verifiable rewards, supervised fine-tuning, and on-policy distillation. Our method, Deletion Tuning, learns a binary mask over the gate projection of transformer blocks. Two practical advantages follow: the trained mask is hundreds to thousands of times smaller than the corresponding FFT checkpoint, and Deletion Tuning forgets far less, which addresses a central challenge in continual learning. Under reinforcement learning, its prior-task accuracy stays near the base level while FFT, LoRA, and random sparse fine-tuning lose up to 5.4 points. To understand why masking alone suffices, we analyze the geometry of the learned weight changes under reinforcement learning using stable rank, spectrum drift, and principal-weight overlap. Compared with LoRA and random sparse fine-tuning, Deletion Tuning is closest to FFT across the three measures.

2611.10866 · ▲ 15 · Code

DepthGraft improves deep language model training by hierarchically sharing key-value representations across layers, reducing memory use while boosting accuracy.

2611.03142 · ▲ 15 · Code

Interactive Visual Gym enables closed-loop evaluation and training of visual interaction models by jointly simulating task environments and human users, exposing limitations in state-of-the-art VLMs and improving performance through closed-loop training.

2611.08286 · ▲ 13 · Code

AAA games offer a rich testbed for evaluating agent perception, reasoning, and control. However, evaluating whether agents can complete tasks in AAA games is challenging and often human-annotated because task outcomes are often difficult to verify automatically, let alone applying it for Reinforcement Learning (RL). We introduce GTAGym, an environment and framework for evaluation and RL on GTA V tasks. Its evaluation benchmark, GTAVBench, derived from GTA V main-story missions, contains 103 online tasks reconstructed across nine task families, alongside 47 image-based shooting and sniping tasks using GTA V imagery. A web server provides access to the online environment and evaluation interface, supporting parallel testing through screenshots and keyboard-mouse actions with automatic engine-state outcome checks. We evaluate six frontier LLMs on GTAVBench with textual task instructions and on its GTAVBench-AGI branch, where instructions appear only in the first screenshot and a harness automatically compresses interaction history. The best configuration achieves 70.9% overall accuracy in GTAVBench-AGI, yet aerial navigation, evasion and interception remain difficult. GTAGym also provides a scalable generator that samples task parameters within user-specified ranges and option sets under fixed templates, creating batches of 3D and image-based shooting environments for RL with automatic reward and termination signals. LoRA RL of Qwen3.8-27B on generated on-foot navigation tasks improves the in-domain performance and transfers to out-of-domain task categories, and native GTA V tasks from 15% to 23%.

2611.13028 · ▲ 12 · Code

Lloyd's k-means algorithm is a long-running standard for clustering, with applications in data analysis, retrieval, quantization and data curation. In this paper, we introduce an exponential moving average optimization procedure for k-means with mini-batch updates, powered by modern deep learning machinery and new insights. We also establish scaling laws on how decay rate and batch size impact final performance. This gives us a formulaic rule to fix them for multiple problem sizes and embedding datasets, without the need for any hyperparameter sweep. The resulting clustering algorithm, coined e-means, is much faster than Lloyd k-means and converges to lower losses, frequently even beating k-means++. On a wide array of embeddings datasets, it beats the loss of Lloyd k-means at 25 epochs in 1.4 to 2.6 epochs, which translates to 10-45x GPU wall-clock speedups compared to state-of-the-art k-means implementations. As an example, e-means partitions DINO10B, a dataset of 10 billion DINO embeddings, into 16 million clusters in only 2.5 epochs, reaching a loss below Lloyd's at 25 epochs.

2611.09501 · ▲ 11 · Code

Agentic workloads are a major source of LLM inference demand. Existing prefill--decode (PD) disaggregated serving systems offer limited flexibility for runtime adaptation of compute allocation and KV cache management, and rely on coarse signals that only partly capture the diverse and dynamic agent workload. These limitations lead to load imbalance between prefill and decode nodes and eviction of KV caches that agents will reuse. We present Evolving Infra, an agent-adaptive serving framework that adapts compute allocation to agent workloads and KV cache management to agent context management at runtime. We decouple compute allocation from key--value (KV) cache management, allowing nodes to switch efficiently between prefill and decode. Building on this design, an online planner matches the PD split to workload composition, while an agent-aware distributed KV manager guides cache eviction and transfer with signals from agent context management. Across coding, math, and computer-use tasks with GLM-5.2 and Kimi K2.6 on up to 256 GB300 GPUs, Evolving Infra reduces mean end-to-end request latency and P95 TTFT by up to 4.1x and 14x over Dynamo, NVIDIA's production framework.

2611.12304 · ▲ 10 · Code

What makes an image creative? Although this may seem like a highly subjective decision, research in psychology suggests that creativity may be based on a set of fundamental creativity dimensions such as novelty, surprise, and arousal. A deeper understanding of image creativity can benefit both research and practical applications. Exploring how human viewers use creativity dimensions to evaluate image creativity may allow us to predict a target image's human-rated creativity automatically. We present a novel framework to estimate the perceived creativity of an image by: (1) predicting the strengths of an image's novelty, surprise, and arousal creativity dimensions, and (2) combining the creativity dimension strengths to estimate an image creativity score. Controlled human-subject experiments were conducted on Amazon Mechanical Turk to collect data to train supervised machine learning models to estimate an image's strengths on the three creativity dimensions. A final machine learning model takes the predicted dimension strengths as input to produce an estimated image creativity score. Results showed that our estimated creativity scores are statistically equivalent to human ratings of creativity for the images we tested. The novel contributions of our research are: (1) identification and validation of three psychological creativity dimensions used to estimate image creativity; (2) supervised machine learning models to automatically estimate novelty, surprise, and arousal strengths for an input image; (3) a supervised machine learning model to predict a creativity score from novelty, surprise, and arousal strengths for an input image, and (4) experimental methods to confirm the accuracy, advantages, and limitations of our approach.

2611.15044v1

The optical continuum variability of quasars is often described by a damped random walk (DRW) model, with characteristic damping timescales $τ_{\rm DRW}$ of hundreds of days and amplitudes $σ_{\rm DRW}$ of $\mathcal{O}(0.1)$ magnitudes. This model has been widely used as a null hypothesis to assess the significance of periodic signals in quasar light curves in the search for supermassive black hole binaries, the significance of extreme quasar variability like tidal disruption events (TDEs) and changing-look (CL) quasars, or to model time lags in reverberation mapping. However, the DRW model is known to be incomplete, and an incorrect null hypothesis can lead to biased significance estimates. We investigate the impact of such null-hypothesis mis-specification on the detection of periodic signals in synthetic quasar light curves and introduce the generative model Nullasar, capable of learning quasar variability directly from time-series data. We demonstrate that Nullasar, trained with flow matching, can learn delicate statistical properties of stochastic light-curve variability without explicitly assuming any noise model. We further demonstrate that Nullasar can learn correlated distributions of noise parameters and generate new light-curve realizations from a range of deviations from the Gaussian DRW noise model, approximating parametric descriptions of stochastic processes. When used as a null hypothesis for detecting periodic signals in synthetic light curves, Nullasar performs comparably to fitting the correct DRW model to each light curve, and outperforms parametric fitting when the variability follows a more flexible AR(2) model. Finally, we identify practical limitations, including sensitivity to contamination of the training set by non-null signals, and show that projecting the conditioning vector of the model onto its leading principal components largely mitigates it.

2611.15040v1

Skins, cosmetic in-game items, are a primary revenue stream for modern online games, yet production decisions (which skins to make, when to release, how to price) remain driven by intuition and retrospective analytics. We present SkinAdvisor, an AI-driven decision framework for skin production planning, powered by SkinCast, a deployed multi-modal Transformer forecasting 180-day per-day sales trajectories for unreleased skins. SkinCast couples four encoders: a pretrained Informer for player-preference time series, a pretrained Player-Hero-Skin graph embedder for relational priors, and two lightweight Transformers for skin attributes and release context, fused through a shared backbone with a local-plus-global day-x positional encoding and decoded by an N-layer Informer in a single non-autoregressive pass. SkinAdvisor plugs these forecasts into a serving pipeline (Select -> Fill -> Infer -> Sales -> ROI -> Ranking) that ranks candidate skins by forecast-based ROI and supports multi-candidate comparison. The system has been in production for a full release cycle. Offline experiments show SkinCast reduces RMSE by ~20% over the strongest baseline (Informer) and ~60% over classical methods; stress tests on a 16-core CPU server confirm p99 latency under 8ms and peak throughput above 1100 rps. We report two deployment lessons: order-of-magnitude accuracy suffices for production decisions, and pretrained embeddings can leak temporal information, requiring season-based splitting to prevent.

2611.15023v1

Multiagent research inherited its economic vocabulary of auctions, markets, prices, bargaining, and matching from a single observed case: human societies. Those institutions bundle identity, competence, execution, ownership, commitment, and settlement in ways that suit agents whose minds are private, who learn rather than copy skills, and who each persist as one person. Engineered agents need not share any of these constraints, making the list of familiar institutions look less like the entirety of economic organization and more like a sample from a larger, unmapped space. We sketch one way to make that space concrete: a provisional chart that records each institution, separately from the agents that run it, as a typed structure over positions, rights and duties, allocation, information, time, state, process, value, and external recognition. In this chart, named institutions become overlapping regions rather than fixed points, neighboring families separate along a few coordinates, and the gaps between them become places to search. We show what the shift in perspective buys with a simulation in which one technical act, copying an agent's competence, is institutionalized as a new independent supplier, as licensed capacity that creates no new participant, and as the splitting of one economic position into successors that inherit some obligations and not others. In a paired task, all three produce an identical technical trace despite differing in who exists, who owns the copy, who is obligated, and who gets paid. We close with a research program: building a principled chart of institution space, mapping how its implementable region shifts with agent capacities, and revising the chart wherever engineered agents expose distinctions it cannot express.

2611.15017v1

Mechanistic localization does not by itself yield usable control: causal sensitivity may occupy little activation energy, an accurate local direction may fail at finite scale, and a reachable numeral target may be wrong. We introduce AROMA (Adaptive Representation-Oriented Multimodal Actuation), a frozen-VLM control framework separating where to intervene, which geometry and gain to use, how to reach a target, which target to pursue, and when to act. In Llama-3.2-11B-Vision-Instruct, four dimensions of a causally validated cross-attention head explain 89.82% of cardinality sensitivity but only 9.52% of median activation energy; magnitude-normalized causal subspace actuation converts this mismatch into finite control, improving exact-count accuracy on untouched procedural data by +8.00 pp. Iterative Minimal-Effort Actuation (MEA) reaches 81.25% of oracle targets versus 15.18% for a frozen direction under matched step budgets. With target and gate frozen before evaluation, targeted MEA improves the previous frozen policy from 34.444% to 38.730% on a new natural Stage-2 cohort (N=630; paired p=4.63e-7) and yields 65.741% versus 65.153% over all 7,992 examples while editing only 63 inputs. Prospective routing, replication across heads and models, and comparisons with attribution-steering baselines further characterize each stage, separating localization, actuation, reachability, target quality, and selective deployment as distinct control problems.

2611.15041v1

Machine learning models are often trained on labeled data produced by multiple human annotators. When those annotators disagree, their labels are commonly collapsed into a single majority vote, which discards information about the underlying disagreement. We examine the cost of this practice for cyberbullying severity, a subjective judgment on which annotators disagree heavily. In the corpus we study, all 5 annotators assigned the same severity label to only 4.1% of comments. Turning 5 judgments into one label hides this disagreement and obscures how much the annotators differed in their assessments. In this paper, we ask whether a model can preserve this information by predicting the distribution of annotator judgments instead of a single majority label. We find 3 main results. First, accuracy against the majority label cannot tell competent models apart: 2 models with similar accuracy can differ greatly in how well they match the distribution of annotator judgments. Second, predicting the full distribution requires little extra effort. Calibrating an existing model, which adjusts its output probabilities after training, captures most of the improvement without retraining, across 3 datasets covering severity, abuse, and irony, while training on label distributions adds a small further gain. Third, we examine a dataset that includes annotator demographics, where a single shared model has one advantage and one limit. The advantage is that a model predicting the distribution of judgments matches each demographic group's judgments significantly better than a model trained on majority labels, so both groups gain. The limit is that it does not narrow the gap in performance between the groups, because one shared model cannot encode which group is right on a given comment. Preserving annotator disagreement is therefore a low-cost step toward fairer moderation, though not a complete one.

2611.15010v1

Failure attribution in multi-agent systems requires tracing error formation to identify the decisive error step and the responsible agent. However, agents' division of work and information exchange lead to Interleaved Steps and Cluttered Content, requiring models to link executions and check details in lengthy records. This makes it difficult to balance localization accuracy and token cost. We find that multi-agent collaboration also leaves records of how executions connect and content passes between them as results are formed, providing evidence for tracing error formation. Based on this observation, we propose CoLFA (Collaboration-Guided Lightweight Failure Attribution), which uses collaboration cues to organize the attribution input according to how results are formed, determining which steps are examined and how their content is presented. Through Handoff-Guided Step Expansion and Cross-Referenced Content Organization, CoLFA helps the model focus on examining error formation and the role of each decision while reducing the processing of irrelevant and repeated records. Across five settings spanning Who&When and LongRCA and three backbone models, CoLFA achieves the highest exact step-localization accuracy with the lowest token consumption. Compared with individual baselines in each setting, CoLFA improves accuracy by up to 37.50% and reduces token consumption by up to 97.6%.

2611.14980v1

Medical vision-language models (VLMs) are compared mainly by accuracy averaged across findings, a summary that can hide complete failure on a single diagnosis. We audit five open medical VLMs (MedGemma-4B, Lingshu-7B, HuatuoGPT-Vision-7B, Qwen2.5-VL-7B, and MedVLM-R1) on 6,048 ChestX-ray14 questions balanced by sex, age, finding, and label. On the 432 pneumothorax-positive X-rays, HuatuoGPT-Vision and Qwen2.5-VL detect zero cases while MedGemma and Lingshu detect 38% and 67% of the same images. The two failing models answer "no" to all 864 pneumothorax cases (balanced accuracy 0.500, chance), yet produce fluent, anatomically specific explanations that name the signs of a collapsed lung and then deny them, with no hedge or abstention. This failure is not an artifact of prompting, image resolution, label ambiguity, or model scale. It persists across four prompt formats, at native resolution, and on single-label cases. A within-model control shows the same forced prompt gives HuatuoGPT 79% recall on nodule but only 6% on pneumothorax. It survives a tenfold scale increase (Qwen2.5-VL-72B detects 4%) and replicates on SIIM-ACR, an independent dataset with radiologist-drawn masks, where the failing models detect 0 of 432 confirmed pneumothoraces, including 0 of the 144 largest lesions. Reading the models' logits and internal states shows this is not blindness. The two failing models are highly confident yet no better than chance and a linear probe recovers pneumothorax from their hidden representations even where recall is zero, so the finding is represented internally but suppressed at the output. Aggregate accuracy therefore cannot establish that a medical VLM can see a given finding, because a model that misses one can still describe it convincingly. Reporting per-finding recall and specificity on balanced cases is needed to tell the difference.

2611.14975v1

The visual world is cluttered: robust vision-language reasoning requires ignoring irrelevant information. Yet vision-language models (VLMs) can suffer substantial accuracy losses when answering questions about a target image with distractors present. Existing mitigations focus on selecting relevant images or limiting information exchange between images. Through a systematic analysis of three open 7-8B VLMs on five perception benchmarks adapted to include distractor images, we show that distractors reduce accuracy even when models correctly identify the relevant image, suggesting that selection alone is not enough. To understand the mechanism behind this interference, we combine attention analysis with targeted knockouts in Qwen2.5-VL and identify middle-layer distractor-to-text connections as a pathway through which irrelevant visual information affects reasoning. We introduce answer consistency regularization, which encourages consistent answer distributions with and without distractors. Without directly constraining attention, this objective reduces middle-layer attention from question tokens to distractor images, consistent with the pathway our analysis implicates. With five distractors, our approach improves circular accuracy by an average of 16 percentage points over the Qwen2.5-VL baseline across the five benchmarks while preserving single-image performance.

2611.15039v1

Deep-learning-based 3D MRI reconstruction is limited in practice by the GPU memory and compute cost of full-volume training; 3D acquisitions are thus most often reconstructed by 2D models applied slice-by-slice and then stacked, which ignores through-plane (stacking-axis) structure. We reconstruct in full 3D and ask whether a model can be trained at low resolution (fast and memory-feasible) yet deployed at higher resolution. This coarse-to-fine strategy needs reliable cross-resolution performance, but standard networks are resolution-specific and degrade substantially when evaluated outside the training setting. We therefore use resolution-agnostic neural operators (NOs) and propose KIR-NO (k-space-to-Image Reconstruction Neural Operator), a discretization-agnostic framework for accelerated 3D MRI reconstruction. KIR-NO is trained with full backpropagation on low-resolution volumes and applied zero-shot to higher-resolution volumes at inference. Its core building block is a 3D neural operator based on discrete-continuous convolutions: convolutional filters are parameterized in a continuous basis and sampled at arbitrary voxel resolutions while preserving the local inductive bias of convolution. KIR-NO integrates these learned operators with physics-based consistency: a k-space neural-operator head first refines the undersampled Fourier measurements, and image-space neural-operator cascades then refine the reconstruction through interleaved data-consistency updates. Across 3D knee and brain MRI benchmarks, KIR-NO improves over classical and CNN baselines: it improves over a state-of-the-art 3D CNN by 2.2 dB PSNR at half resolution under 4x acceleration on SKM-TEA, with gains persisting at 8x and 16x. These results show that resolution-agnostic operators make coarse-to-fine 3D MRI reconstruction - cheap low-resolution training with high-resolution deployment - practical across anatomies.

2611.14979v1