TodayNewsResearchFieldsStacksMarketAnalytics
⌘K

Research

New papers ranked by attention, with their code, models and discussion linked. Pick a field on the left.

WeekMonthQuarterAll
NewestMost relevant

Papers

Open problems

Rising research topics

Ask the research

Sample
ByteTrail · AI and developer-tools intelligence
NewsResearchStacksAPI
Fields · this week
All fields3148Multimodal507Agents486Frontier LLMs421Inference & serving254Image segmentation1703D scene153Learning verifiable rewards146Graph learning142Voice AI94Classifier & causal discovery83Forecasting time74Flow matching68Evals65Federated learning48Teacher policy distillation47All 32 fields

Open problems across fields

  • General in-context learning for embodied agents1 paper raises it · 0 address it

    Enables models to adapt to new tasks and environments without retraining, improving robustness in real-world deployment.

    • Has GPT-6-Astra Achieved Embodied Intelligence?raises
  • Cost and latency in embodied control loops1 paper raises it · 0 address it

    High computational and time costs limit real-time applicability of agentic systems in dynamic environments.

    • Has GPT-6-Astra Achieved Embodied Intelligence?raises
  • Physical-state verification and failure recovery1 paper raises it · 0 address it

    Agents must reliably detect and correct errors in physical interactions to avoid irreversible failures in real-world tasks.

    • Has GPT-6-Astra Achieved Embodied Intelligence?raises
  • Long-horizon task preservation in embodied agents1 paper raises it · 0 address it

    Maintaining progress over extended sequences is critical for complex, multi-step tasks in robotics and navigation.

    • Has GPT-6-Astra Achieved Embodied Intelligence?raises
  • Open-source stacks for generalist computer-use agents1 paper raises it · 1 addresses it

    Closed systems hinder reproducibility and systematic study of agent design for real-world computer tasks.

    • OpenGCUA: Toward a Fully Open Stack for Generalist Computer Useraises · addresses

Named by an AI from these papers' abstracts.

Rising research topics

  • Training & fine-tuning44+120.0%
  • Computer vision33+83.3%
  • AI safety30+76.5%
  • Evals65+51.2%
  • Code models14+40.0%
  • Federated learning48+29.7%
  • Image & video generation22+29.4%
  • Robotics18+28.6%

Papers

  • arXiv cs.MAOct 16

    The Weight of Teamwork: Does Communication Structure Help LLM Scientists Search for Scientific Laws?

    Jianing Shi, Zaibin Zhang, Shijun Lei, Zhenfei Yin

    Team performance is often reported as one delivered score, even though that score is produced in two stages: a system generates candidate answers and a selection rule returns one of them. We use symbolic regression as an observable scientific-search setting and compare fixed-budget LLM teams that differ in peer visibility and update timing, evaluating both their frozen candidate archives and the expressions selected from those archives. Distributing the call budget across more independent trajectories produced the largest measured gain, improving mean held-out fitting score Q by 0.70 at the widest tested allocation. Under a common archive-wide selector, the implemented Ring policy reduced mean Q by 0.33 relative to Independent, while Star and Complete remained unresolved. Holding the generated archives fixed, changing the delivery rule changed the measured architecture gaps, including the sign of the Complete point estimate. These results support reporting candidate generation and final selection as separate endpoints. They do not identify a communication mechanism or a universally preferred topology; rather, they show that an architecture ranking is not identified by…

    PDF
  • arXiv q-bio.GNOct 16

    Reducing double-dipping in single-cell topic models with three-way count splitting

    Zheng Qu

    Topic models learn gene programs from single-cell RNA sequencing (scRNA-seq) data, and the same counts are then used to test which genes and donors are associated with those programs. Count splitting is the standard remedy for this reuse, but it has not been tested when an encoder reads each cell's counts. On a 104,400-cell lupus cohort with 300 permuted null genes per run, naive tests rejected 36% of null gene-topic pairs at a nominal 5% level, and the usual two-way split rejected 38%, because the encoder sees the tested gene's own held-out count. Encoding topic usage on a separate third of the counts, with beta-binomial thinning, lowered this to 7.7% (6.7% for genes outside the model); for two other models, LDVAE and scETM, the rate was 7.0-7.3% (naive 51-59%). A label-free interferon program shifts in held-out counts in the lupus cohort and in a pre-specified COVID-19 cohort, and donor-level topic usage predicts disease (AUROC 0.96) beyond cell-type proportions (0.86 and 0.58 in the two cohorts). Calibration and prediction did not depend on the model. We do not claim better clustering, prediction or topic stability. Code, figures and result tables are public (MIT license; DOI…

    PDF
  • arXiv cs.CROct 16

    Recording-Aware Evaluation of Multimodal OT Intrusion Detection

    Mustafa Umut Ozbek, Khalil El-Khatib, Li Yang

    Machine-learning-based intrusion detection systems (IDSs) for operational technology (OT) are commonly evaluated on benchmark datasets in which normal and attack data are collected in separate recordings. In such settings, high performance may reflect recording-associated differences as well as attack behavior. We introduce a recording-aware evaluation procedure that compares attack windows with background from the same recording, uses prior data for model fitting and calibration, tests temporal alignment against randomized event schedules, and reports event coverage together with alarm burden. On the ProvICS multimodal dataset, cyber modalities strongly distinguish the benign and attack recordings before the first labeled attack. Same-recording evaluation reverses the apparent modality ranking across all studied window sizes: the physical view becomes the strongest individual modality, while cyber-only fusion shows the largest difference between cross-recording and same-recording AUC, reaching 0.54. Operational analysis further shows that permissive event-overlap metrics can overstate detection when a small number of long alarm episodes cover multiple actions. An external…

    PDF
  • arXiv cs.LGOct 16

    Data-Driven Resource Aggregation for Scalable Digital Twin and Simulation Model Generation

    Yucheng Wu, Alp Akcay

    As industrial digitization accelerates, the use of digital twins has become widespread. However, modern manufacturing systems are highly dynamic and frequently reconfigured, making manual model development and updates increasingly difficult. Recent research on automated simulation model generation shows that historical data such as event logs can yield simulation models automatically. However, modeling every resource at full granularity leads to unnecessary complexity and computational cost. We propose a method for automatically selecting and aggregating resources to simplify simulation modeling. Our method uses ideas from inverse reinforcement learning to learn behavioral value representations of resources from production event logs. This process identifies and aggregates resources with similar functional contributions. Experiments on manufacturing systems, using throughput and system time as performance measures, demonstrate that value-based resource aggregation performs well when the goal is to substantially reduce complexity while preserving performance accuracy.

    PDF
  • arXiv cs.LGOct 16

    On the Selectivity of Generative Models in Structure-Based Drug Design

    Ella Miray Rajaonson, Jungyoon Lee, William Chau, Alan Aspuru-Guzik, Benjamin Sanchez-Lengeling +3

    Designing small-molecule drugs that selectively bind to a target protein while avoiding unintended interactions with other proteins is essential for reducing side effects that contribute substantially to clinical attrition. Although structure-based drug design (SBDD) has advanced rapidly with deep learning, most generative models still optimize binding to a single target and disregard selectivity due to limited training data and unknown off-target interactions. Motivated by this gap, we introduce SELECTBENCH, a comprehensive benchmark assessing the selectivity of generative SBDD models. SELECTBENCH provides three curated evaluation tasks reflecting different real-world drug discovery scenarios: literature-based case studies, safety screening panels used in pharmaceutical development, and large-scale structure-based evaluation sets. We further introduce a suite of metrics to quantify selectivity and conduct a systematic evaluation of nine representative SBDD models, including methods explicitly trained for selectivity and widely-used inference-time guidance techniques. Our results show that even selectivity-trained models achieve only modest off-target discrimination and…

    PDF
  • arXiv cs.CLOct 16

    Prefix-Tuning for Arbitrary Output Sequences on Pretrained Transformers

    Peter Cho-Ho Lam, Ziyi Wang, Zirui Zhou

    Prefix-tuning is a parameter-efficient fine-tuning method for large language models (LLMs) that steers model outputs by prepending task-specific prefix representations to the input sequence. Despite its practical success, the theoretical understanding of the expressiveness of prefixes remains incomplete. In particular, the following question remains open: Given a pretrained autoregressive Transformer, for any input-output sequence pair $(\mathtt{x},\mathtt{y})$, does there always exist a prefix $\mathcal{S}$ such that prepending $\mathcal{S}$ to $\mathtt{x}$ induces the model to generate the target output $\mathtt{y}$? In this work, we prove that under the assumption that $\mathtt{y}$ is in the range of the pretrained Transformer, such a prefix $\mathcal{S}$ exists and can be explicitly computed by solving multiple linear systems. We further show that this assumption is necessary, in the sense that such a prefix may not exist when the assumption fails, thereby providing a complete answer to the above question. Our analysis is constructive and the obtained theoretical results hold under fairly general assumptions on model architecture. Extensive experiments on GPT-2 and Qwen3-4B…

    PDF
  • arXiv cs.CVOct 16

    HAVANA: A Hierarchical Video-Action Model for Vision-Language Navigation

    Dongzhihan Wang, Yi Du, Chen Wang, Xiuxian Li, Liang Xu

    Vision-Language Navigation (VLN) requires long-horizon instruction-consistent planning and short-horizon local planning in current geometry. Video-action models (VAMs) offer a promising way to learn predictive representations for navigation. However, existing VAMs jointly predict visual futures and local trajectories in a single process, even though they operate at different paces: future prediction remains stable across multiple updates while local trajectories must react to fresh observations. To bridge this gap, we introduce HAVANA, a dual-system VAM that decouples predictive reasoning from reactive local planning. A slow system produces a compact language-conditioned latent plan from observation history, while a fast system combines the latest plan with the current observation to generate local trajectories. Their asynchronous execution enables rapid adaptation without repeated future prediction, and no future video is generated during inference. Experiments on OmniNavBench demonstrate state-of-the-art navigation performance over prior baselines, while real-world deployments further show robust generalization to unseen environments across different robot embodiments. Our code…

    PDF
  • arXiv cs.CLOct 16

    Do LMs Know What Variables Do? Interpreting Code Representations via Variable Roles

    Célian Vasson, Daniil Ognev, Kentaro Inui, Benjamin Heinzerling

    Variable roles describe recurring behaviors such as accumulation, traversal, and retaining the best candidate encountered. We use Sajaniemi's independently defined taxonomy to study which role distinctions can be recovered from large language models, how sparse autoencoders (SAEs) retain those distinctions, and whether associated directions support control of generation. We introduce static annotation rules for all eleven roles and a corpus of 126,926 natural-code files across five languages. Across five model-SAE systems, residual probes outperform individual SAE features and tested lexical baselines, with modest losses after neutral target renaming. Combining SAE features closes much of the single-feature gap, but probes on SAE reconstructions are less sensitive to role-changing edits reviewed by LLMs. High paired ranking accuracy coexists with frequent false positives at fixed thresholds. During generation, interventions often change role annotations at the cost of correctness; changes accompanied by passing tests show no clear aggregate advantage over random controls. The benchmark separates role-related recoverability, SAE fidelity, and correctness-preserving control through…

    PDF
  • arXiv cs.LGOct 16

    Graph-Regularized Synthetic Control for Stable Counterfactual Estimation under Donor Redundancy

    Yanlin Zhang, Sungyong Chung, Alireza Talebpour

    Synthetic control methods estimate counterfactual outcomes by weighting untreated donor units to match a treated unit before treatment. In recent applications, however, the donor pool can be large, highly correlated, and short in pre-treatment history, yielding many near-equivalent solutions with unstable and difficult-to-interpret donor weights. We propose Graph-Regularized Synthetic Control (Graph-SC), a convex estimator that augments the classical pre-treatment matching objective with a graph Laplacian penalty. The graph encodes similarity among donor trajectories, encouraging coherent weights over structurally related donors while preserving the transparency of synthetic controls. We provide theoretical results characterizing the bias-variance tradeoff induced by graph regularization and establish efficient convergence guarantees with an accelerated proximal gradient method. Across a controlled simulation under donor redundancy, the California Proposition 99 benchmark, and observational counterfactual analyses on the Waymo Open Motion Dataset, Graph-SC improves weight stability, reduces sensitivity to redundant donor pools, and maintains competitive pre- and post-treatment…

    PDF
  • arXiv cs.LGOct 16

    Looped GNNs Beyond Fixed Depth: Graph Size Generalization and Parameter-Efficient Adaptation

    Zhiyang Wang, Zhishang Luo, Yusu Wang

    Generalizing from small graphs to larger ones can require computation beyond the depth of a neural model used during training. We study target tasks that admit a common local iterative computation and whose effective dependency radius grows with graph size. The central idea is to learn a local update rule through GNN pretraining, then repeatedly apply the same learned processor beyond the training depth to extend information reach on larger graphs without retraining. Calling such parameter-shared architecture a \emph{looped GNN}, we categorize six schemes for looped GNNs to separate the effects of depth extension, fine-tuning, and freeze-and-steer adaptation. Our analysis establishes fixed-depth information limits and sufficient conditions under which accurate iterative reuse overcomes them. To improve beyond unadapted reuse, we introduce \emph{Loop-Freeze-Steer} (LFS), which freezes the pretrained processor and learns a small controller to steer its repeated application, using much fewer trainable parameters than full fine-tuning. Our finite-sample analysis gives sufficient conditions for adaptation to improve performance on larger graphs. Experiments on synthetic tasks and chip…

    PDF
  • arXiv physics.ao-phOct 16

    Interpretability of AI weather models via intermediate decoding

    Matthias Beylich, Kirsten I. Tempest, Tobias Selz, George C. Craig

    AI weather prediction models are often regarded as black boxes. Despite their impressive skill, there is little understanding of how the input data is processed during inference to produce the forecast. This limitation arises from the use of high-dimensional latent representations in which individual dimensions are no longer directly interpretable in physical terms. To gain insight into how forecasts are constructed, we propose the method of intermediate decoding for AI weather models and apply it to GraphCast, Aurora and ArchesWeather. In this approach, latent states from intermediate depths are decoded back into meteorological fields. To do this, we use the decoder that has been trained to output the forecast at the end of processing. Since latent states at different depths do not share a consistent basis, we train affine transformations that align intermediate latent states with the decoder's expected input space. Applying this framework to the three models yields the following results. First, the skill of forecasts at intermediate stages of processing improves gradually through the processor layers. Second, bypassing all processor blocks shows that the encoder-decoder pair is…

    PDF
  • arXiv cs.CVOct 16

    BLANC: Adaptive Temporal Scheduling for Sign Language Production

    Youngmin Kim, Seoyeong Hwang, Sujung Hong, Seong Jae Hwang

    Sign language production (SLP) aims to synthesize expressive signing motion from written text. Modeling signing motion presents two key temporal challenges. First, fine-grained hand articulations demand higher temporal resolution than relatively slower body movements. Second, signing events unfold over variable durations, calling for adaptive temporal scheduling during generation. To address these challenges, we first introduce a Multi-Rate Vector Quantizer (MRVQ), an articulation-decoupled motion tokenizer that represents the body and both hands as separate discrete streams at distinct temporal resolutions. Building on this representation, we propose BLANC (BLock-wise Adaptive diffusioN with Articulatory Clock), a gloss-free SLP framework that generates variable-duration motion blocks causally while denoising tokens in parallel within each block. At its core, a Latent Articulatory Clock (LAC) learns temporal structure from multi-scale motion-derived partitions to predict block durations and sequence termination during generation. We further introduce inference-time body-conditioned hand refinement to improve coordination across articulators. Extensive experiments on…

    PDF
  • arXiv cs.AIOct 16

    Joint, Decoupled, or Agent: When Each Multimodal Alignment Paradigm Wins

    Yi Fang, Gaurav Srivastava, Sirui Ding, Xuan Wang

    Jointly pretrained multimodal encoders are a common way to combine modalities, while alternatives integrate their information during downstream training or inference. When does each approach win? We compare three paradigms, instantiated with public checkpoints, across 16 paired datasets in five domains and multiple training-label budgets. JOINT trains a fusion head on frozen, jointly pretrained encoders. DECOUPLED trains a fusion head on frozen, independently pretrained encoders. AGENT gives a frozen LLM task instructions, labelled examples, available sample text, and supervised unimodal readouts. No paradigm dominates the complete 144 dataset-budget combinations: DECOUPLED wins 54, JOINT 44, and AGENT 46 by mean utility. ScienceQA, MELD, and MetMeme shift from AGENT to fusion as supervision increases. Between the fusion paradigms, JOINT leads on four of five QA datasets across all evaluated budgets, while higher unimodal 5-NN accuracy is associated with a relative shift toward DECOUPLED across eight classification datasets (Spearman $ρ=-0.67$). We test whether representation and task diagnostics predict these preferences on held-out datasets before fitting their candidate…

    PDF
  • arXiv cs.AIOct 16

    Cache2LoRA: Understanding and Improving Context Internalization

    Yi Fang, Gaurav Srivastava, Zhenyu Bi, Sirui Ding, Xuan Wang

    Context-to-LoRA methods encode documents as low-rank parameter updates, allowing language models to answer without retaining the source text in their decoding context. How reliably do these updates preserve facts and support their use under different instructions? Controlled diagnosis of existing context-to-LoRA methods separates fact discrimination, target retention, and response adaptation. Competing sources influence answers, while target-only construction and memory-state optimization improve recovery at fixed update size. Recovering facts also does not ensure that answers follow response instructions. We introduce Cache2LoRA, which uses each query to aggregate evidence from a reusable frozen document cache before generating LoRA updates. Trainable encoding operates only on the query and memory-token suffix, and multi-style supervision teaches both evidence preservation and answer expression. Across 17 tasks, Cache2LoRA improves on existing context-to-LoRA methods and achieves macro F1 of 0.663, approaching full-context in-context learning at 0.671. For 32K documents and 256-token outputs at batch size 16, its decoding throughput is 5.9$\times$ that of full-context ICL. At…

    PDF
  • arXiv cs.CVOct 16

    Hierarchical and Query-Aware Visual Token Pruning for High-Resolution LVLMs

    Luther Wei, Ling Zou, Kang-Yang Huang, Yu-Syuan Xu, Wen-Huang Cheng

    High-resolution Large Vision-Language Models (LVLMs) preserve critical, fine-grained visual details by encoding a global thumbnail alongside multiple local crops. However, the resulting thousands of vision tokens introduce severe computational bottlenecks during Large Language Model (LLM) inference. Existing token pruning methods rely primarily on query-agnostic visual saliency, while current query-aware approaches either operate at a coarse crop-level resolution or compute token relevance too late, after vision tokens have already entered the heavy LLM layers. To address this, we propose a training-free, hierarchical token pruning framework that utilizes query guidance prior to executing any LLM blocks. Our top-down hierarchy first operates at the instance-level, dynamically scaling the vision token retention ratio of each sample based on its visual complexity. Next, at the crop-level, we distribute this sample budget across local crops using sentence-level cross-modal similarity. Finally, at the token-level, we exploit projected vision tokens alongside textual query token embeddings in the shared space to measure fine-grained alignment, applying a text-debiasing normalization…

    PDF
  • arXiv cs.CVOct 16

    BiNar: Bitstream Native Restorer for Bitstream-corrupted Image Restoration

    Fangcheng Li, Zhen Yu, Kejun Wu, Tianyi Liu

    Bitstream-corrupted image restoration (BCIR) aims to reconstruct faithful images from corrupted compressed bitstreams. However, corruption occurs in bitstream-domain may lead to unpredictable error propagation and distortion amplification in pixel-domain, and even result in complete undecodability by standard decoders, posing significant challenge to existing image restoration paradigms. To address this challenge, we propose Bitstream Native Restorer (BiNar), a generative restoration framework that leverages fine-grained bitstream-native semantics to guide bitstream-corrupted image restoration. In BiNar, Bitstream Semantic Modeling (BSeM) extracts robust informative semantics and models sequential bytes directly from corrupted bitstream, enabling to generate fine-grained bitstream descriptions without relying on pixel decoding. Based on the descriptions,the Corruption-adaptive Multimodal Semantic Aggregation (CMSA) further extracts adaptive visual modal cues in decoded corrupted images and aggregates to bitstream descriptions. The multimodal aggregated fine-grained descriptions can serve as coherent restoration guidance. Finally, Blind Corruption Localization (BCoL) exploits the…

    PDF
  • arXiv cs.SEOct 16

    Does AI-Assisted Development Homogenize Software? An Empirical Study of Code Similarity, Quality, and Evolution

    Luciano Marchezan, Romain Lefeuvre, Kevin Delcourt, Benoit Combemale, Houari Sahraoui

    AI-assisted software development is evolving from interactive LLM assistance toward agentic systems that autonomously perform software engineering tasks, yet the effects of this shift remain insufficiently understood. We present an empirical study of 50 long-lived Java and Python open-source projects across three eras: pre-LLM, potentially LLM-assisted, and detected agentic development. Using snapshots from Software Heritage, we examine code similarity, software quality, and software evolution. Similarity is measured semantically through embedding-based cosine similarity and syntactically through CodeBLEU*. Quality is assessed with SonarQube metrics covering maintainability, complexity, reliability, and security. Evolution is characterized by annualized growth rates of LOC, functions, classes, files, and statements, and by commit activity in the three months preceding each snapshot. Results show no consistent evidence of increased similarity as semantic and syntactic similarity decreased significantly in Java for both the LLM and agentic eras relative to the pre-LLM era, while Python showed no significant differences. Software quality remained largely stable, although the agentic…

    PDF
  • arXiv cs.AIOct 16

    Persistent Memory and Iterative Improvement in an LLM Agent for Chinese-Character Drawing

    Peilin Wu, Yi-Zhuang You

    Does persistent memory accumulated from previous attempts help an LLM agent solve new problems, and does memory design affect that benefit? We study programmatic Chinese-character drawing across five agent conditions attempting the same 668 items. Accumulated memory produces modest aggregate gains despite substantial growth in the memory stores: free-form notes improve first-attempt success by 4.6 percentage points, while a design combining free-form memory with reusable code improves it by 2.8 points. The two designs do not differ significantly overall, but the code bank shows a suggestive advantage on later, harder items. A separate post-hoc comparison finds a much larger improvement when the code-bank agent receives privileged, target-specific structural supervision. The agent can therefore exploit highly informative structural guidance, although the privileged per-item information does not establish what could have been learned from experience alone. These results suggest that memory design may affect where accumulated experience becomes useful, even when its average benefit is modest.

    PDF
  • arXiv cs.CLOct 16▲ 291

    OR-License: Evaluating LLM Agents as Real-World Operations Research Practitioners

    Peisong Wang, Feixue Liu, Yize Meng, Bowen Liu, Ruoyu Chen +14

    Can large language model (LLM) agents deliver expert-level operations research (OR) solutions directly from original client materials, without prior expert reformulation? Existing OR benchmarks typically begin with self-contained descriptions or predefined optimization tasks, underrepresenting the translation of original business materials into professionally validated solutions. We introduce OR-License, an end-to-end benchmark of 31 instances across 10 industrial scenarios from paid OR engagements totaling USD 3.68 million. It preserves original requirements, input data, deliverable formats, and information boundaries after de-identification. To study information acquisition, we introduce ORSim, a controlled client simulator that discloses business facts without supplying OR expertise. Hidden evaluation checks feasibility and compares objective quality against original expert solutions accepted by clients before benchmark construction. Deployment-Adjusted Quality (DAQ) averages expert-relative quality across instances, assigning zero to infeasible deliveries. Across 11 frontier LLMs using coding-agent harnesses, several exceed the expert reference in conditional objective…

    Code
  • arXiv cs.GTOct 16

    Coopshift: Post-Training Amplifies The Memory Curse In LLM Cooperation

    Jiayuan Liu, Shiyi Du, Tianqin Li, Emanuel Tewolde, Keyang Xuan +1

    How does post-training change cooperation during repeated interaction? Within each of seven released model lineages, we compare pretrained (Base), supervised fine-tuned (SFT), and subsequent preference- or reward-optimized (Aligned) checkpoints across seven social dilemmas. Copies of each checkpoint play one another for 200 rounds, with visible history limited to the last two or forty rounds. The memory curse refers to lower cooperation when agents can see more past rounds. Our central finding is that post-training amplifies this decline. Averaged across the six binary-action games, expanding visible history reduces cooperation at every SFT and Aligned checkpoint. At each post-trained stage, these declines exceed Base's in six of seven lineages. Most of this average amplification is already present at SFT. With long history, both post-trained stages also cooperate less than Base on average in each binary-action game. In the symmetric binary games, their fitted cooperation rates fall more steeply on average than Base's as per-round cooperation costs rise, benefits fall, or noncooperation becomes more rewarding. The difference in cooperation rates between Aligned and SFT also…

    PDF
Load more
PDF