Research Paper Cockpit

Daily Digest - 2026-11-02

Papers first seen in this daily snapshot.

Daily Archives

Quick jump into generated daily digests.

Research Workflow

Latest digest: 2026-11-17.

Papers

51 visible entries

arxiv Score 27.3

Latent Reasoning in Continuous Space for Unified Multimodal Models

2026-10-30 · Byungwoo Jeon, Yoonwoo Jeong, Hyunseok Lee, Minsu Cho, Jinwoo Shin

General AI

Unified Multimodal Models (UMMs) achieve remarkable success on diverse multimodal understanding and generation tasks, yet they still struggle to effectively leverage language-based chain-of-thought reasoning for image generation. As a result, existing reasoning-augmented UMMs often fail to inherit the test-time scaling…

Review
pending
Role
unreviewed
Read
now
arxiv Score 21.3

ImplicitHop: Evaluating knowledge integration across context and memory

2026-10-30 · Alexander Detkov, Dashiel Karish, Matthew Oleynikov, Matt Thomson

Research Track A · General AI

The ability to integrate new information and deduce its consequences is central to learning, understanding, and reasoning in large language models (LLMs). Whether information arrives during training or in-context, models must not only store local facts, but propagate their long-range implications globally. To explore i…

Review
pending
Role
unreviewed
Read
now
huggingface Score 19.5

SketchDirector: Complex Scene Sketching via Agentic Layout Reasoning and Detail Refinement

2026-10-30 · Yiqi Liang, Xiaodong Cun

General AI

Existing sketch generation methods often rely on image guidance and are largely limited to individual objects or simplified scenes, leaving text-guided complex scene sketching underexplored. We present SketchDirector, an agentic framework that mimics the human sketching workflow for automatic multi-concept scene sketch…

Review
pending
Role
unreviewed
Read
now
arxiv Score 19.3

Subquadratic Attention with a Hierarchical Tree-Structured Indexer

2026-10-30 · Leon Lufkin, Cutter Dawes, Alexander Speicher, Kamesh Krishnamurthy, Tomás Figliolia, Beren Millidge

General AI

The quadratic scaling of attention is a fundamental bottleneck limiting the long-context capabilities of large language models (LLMs). A variety of approaches have attempted to circumvent this, but they either fail to match attention's expressiveness (e.g., sliding-window or linear attention variants with fixed-size st…

Review
pending
Role
unreviewed
Read
now
arxiv Score 19.3

When the Past Shapes Action: Understanding Embodied Memory

2026-10-30 · Siqi Li, Ziyi Bai, Xinrun Xu, Tinglei Huang, Börje F. Karlsson

General AI

Recent advances in generative foundation models have expanded the capabilities of embodied AI, enabling increasingly general navigation, manipulation, and long-horizon task execution. However, effective interaction with dynamic environments requires more than only strong perception and reasoning: agents must act with p…

Review
pending
Role
unreviewed
Read
now
arxiv Score 17.3

Aligning MLLMs with the Latent Structure of Human Cognition via Behavior-Derived Semantic Dimensions

2026-10-30 · Ning E, Changde Du, Yizhuo Lu, Huiguang He

General AI

Multimodal Large Language Models (MLLMs) have achieved impressive performance across a wide range of reasoning tasks, yet it remains unclear whether their internal representations and decision criteria align with the latent structure of human cognition. In this work, we propose a novel framework to align MLLMs with the…

Review
pending
Role
unreviewed
Read
now
arxiv Score 17.3

TradeWar: A Benchmark for Execution-Grounded Trade Negotiation with Adaptive LLM Agents

2026-10-30 · Xiangyu Kong, Ruirui Yang, Xia Hou

Research Track A · General AI

Multi-turn negotiation remains a challenging problem for large language model (LLM) agents. While LLMs can conduct natural-language dialogue, existing benchmarks predominantly assess only agreement rates, overlooking whether negotiated commitments actually create value once executed. To address this gap, we introduce T…

Review
pending
Role
unreviewed
Read
now
huggingface Score 15.5

When Requirements Interact: Scaling Long-Horizon Coding via Interdependent Specifications

2026-10-29 · Yishuo Cai, Minzheng Wang, Liang Zhao, Linghao Zhang, Lei Li, Shicheng Li, Xu Sun, Fuli Luo

General AI

To succeed in real-world long-horizon software development, coding agents must satisfy multi-dimensional task specifications, such as interface compatibility and error handling, and do so without breaking existing dependencies. However, current works lack specification in both quantity and diversity, often emphasizing …

Review
pending
Role
unreviewed
Read
now
arxiv Score 15.3

Open-Weight MLLMs Cannot Compare What They Cannot Name

2026-10-30 · Wenxin Su, Anna Kreshuk

Research Track A · General AI

Multimodal Large Language Models (MLLMs) are increasingly capable of understanding images and performing complex visual reasoning, suggesting the possibility of using them to automate biological image analysis. This need is becoming increasingly urgent as advances in technology continue to generate ever larger and more…

Review
pending
Role
unreviewed
Read
now
arxiv Score 15.3

RoboArchitect: Automating the Design of Robotic Agents with Persistent Memory

2026-10-30 · Shengguang Wu, Suning Huang, Cherie Ho, Ria Doshi, Hao Zhu, Serena Yeung-Levy

General AI

Recent work shows that a frozen robot policy performs far better inside an agent harness, where control code decides when to query the policy, consult vision tools, or command the arm directly. Yet today these harnesses are built by hand and redesigned for each new robot or benchmark. We show that harness design can be…

Review
pending
Role
unreviewed
Read
now
huggingface Score 14.5

SciGame: Let Agents Play the Scientific Discovery Games in the Real and Counterfactual World

2026-10-27 · Yusheng Zhao, Bowei He, Xue Liu, Mingxiao Li

General AI

Large language model agents are increasingly used for scientific discovery. However, existing benchmarks mix law inference with programming, score data fitting instead of prediction on unseen inputs, and do not separate empirical evidence from pretrained priors. We introduce SciGame, a controlled benchmark that convert…

Review
pending
Role
unreviewed
Read
now
huggingface Score 14.5

REA: A Routed Embodied Agent Runtime for Asynchronous Brain-Cerebellum Control

2026-10-30 · Pengfei Zhou, Yixing Ma, Zhengfeiyang Zhang, Haoyuan Chen, Zhenglin Wan, Hexin Wang, Wangbo Zhao, Yang You

General AI

Foundation models are becoming remarkably reliable at high-level reasoning, yet they remain too slow and costly to serve as fine-grained controllers. Conversely, fast policies like VLAs and whole-body controllers offer more efficient reflexes, but struggle to generalize. While pairing a deliberative brain (foundation m…

Review
pending
Role
unreviewed
Read
now
arxiv Score 14.5

RetryRoute: Retry-Aware Model Routing for Agentic Image Editing Services

2026-10-30 · Hanshuai Cui, Zhiqing Tang, Zhi Yao, Jun Luo, Weijia Jia

Research Track B · General AI

Web agents increasingly edit images by calling open editing models in a critique-and-retry loop, and these models differ in cost by more than an order of magnitude. After every rejected edit, a service must decide whether to retry the same model, escalate to another or give up. Existing routers pick one model per reque…

Review
pending
Role
unreviewed
Read
now
arxiv Score 14.3

Think in 360°: Benchmarking and Bootstrapping Panoramic Video Reasoning

2026-10-30 · Lidong Lu, Lidong Zou, Xinyi Mao, Jiaxin Lv, Zhenquan Li, Liangrui Dong, Tong Lu, Guo Chen

General AI

A panoramic video captures all directions around the camera, but relevant evidence can appear in different directions and at different times. Understanding it requires models to connect observations across the viewing sphere and time. We introduce PanoVideoBench, the first manually annotated benchmark for comprehensive…

Review
pending
Role
unreviewed
Read
now
arxiv Score 14.0

Lifelong Learning of Scheduling Rules with Surrogate-assisted Genetic Programming for Dynamic Flexible Job Shop Scheduling

2026-10-30 · Luyao Zhu, Fangfang Zhang, Yi Mei, Ke Chen, Mengjie Zhang

Research Track A

Genetic programming (GP) has been widely used to learn scheduling rules for dynamic flexible job shop scheduling (DFJSS). However, trained rules tend to only generalise to tasks of similar scales and distributions, and when task characteristics change, learned rules often suffer significant performance degradation. Mor…

Review
pending
Role
unreviewed
Read
now
arxiv Score 13.5

Certified Behavioral Retention for Continual Adaptation with Held-Out Anchors

2026-10-29 · Ahanaf Hasan Ariq, Abrar Shahid

Research Track A

Continual adaptation requires deciding not only how to update a model, but whether to commit an update without exceeding a declared budget for changes to protected behavior. We introduce Certified Anchor-Based Adaptation (CABA), a candidate-agnostic post-update gate that reserves historical observations from candidate …

Review
pending
Role
unreviewed
Read
now
huggingface Score 13.5

MemSteer: Turning User Histories into Supervision for Agent Memory Optimization

2026-10-29 · Jie Huang, Cuiling Lan, Qize Yu, Xiulian Peng, Yan Lu

General AI

An effective agent memory system should adapt to a user's history and task requirements. Human-annotated supervision can guide this adaptation by providing feedback for evaluating and improving memory designs. Yet suitable labels are rarely available for new user histories, and manually constructing QA pairs and suppor…

Review
pending
Role
unreviewed
Read
now
arxiv Score 13.3

DecisionRank: On-Policy Distillation for Decision-Model Rerankers

2026-10-30 · Yanran Wu, Dian Yu, Chongyang Bai, Jieyi Zhang, Rui Liu, Tianyi Chen, Shike Mei, Hang Cui, Frank Shyu, Sandeep Pandey, Luke Simon, Chunwei Liu, Xi Liu

General AI

Decision models target structured tasks by directly predicting choices or scores, avoiding autoregressive decoding latency. However, transferring a generative LLM's ranking knowledge to a decision model remains challenging: the teacher expresses preferences through a ranked list of document ID sequences, while the stud…

Review
pending
Role
unreviewed
Read
now
arxiv Score 12.5

ANGLER: Verifier-Grounded Benchmarking with an Adaptive Attacker for Web Agents

2026-10-30 · Yini Huang, Aiwen LU, Zhen Sun, Yi Han, Yuzhi Zhao, Jiaheng Wei

Research Track B · General AI

Web agents are increasingly deployed to complete tasks autonomously on real websites, where adversarially injected content can induce them to perform unauthorized actions. Existing evaluations primarily use fixed attacks or focus on a single attack type, and evaluate attacks at the text level rather than through real s…

Review
pending
Role
unreviewed
Read
now
arxiv Score 12.3

Hierarchical World Models with Implicit Dynamics

2026-10-30 · Gaoyue Zhou, Yvonne Wu, Zichen Jeff Cui, Nicolas Ballas, Mido Assran, Lerrel Pinto, Yann LeCun

General AI

Planning with world models enables test-time adaptation for embodied agents, offering superior generalization over direct policy learning. However, long-horizon tasks remain hindered by cumulative errors and search complexity. Existing hierarchical methods often use rigid macro-actions or latent skills, leading to brit…

Review
pending
Role
unreviewed
Read
now
arxiv Score 12.3

KITE: Retrieval as a Decision Procedure for Long-Term Conversational Memory

2026-10-30 · Feng Li, Kailun Tan, Zhibo Zhou, Youxiang Wang, Lejun Liu, Huaigu Li, Fei Du, Huilong An

Research Track A · General AI

Many long-term memory systems for LLM agents are built around embedding-based retrieval. Al- though effective for semantic matching, similarity ranking alone does not establish whether re- trieved memories satisfy the conditions implied by a question. In this paper, we introduce KITE (Knowledge-Indexed Temporal Evidenc…

Review
pending
Role
unreviewed
Read
now
arxiv Score 11.5

AutoHebb: Self-Modifying Neural Networks that Zero-Shot Infer and Continually Learn

2026-10-29 · Daniel Levine, David van Dijk, Kenneth Kay

Research Track A

Humans and animals have a unique capacity to generalize and learn in novel circumstances, including "0-shot" inference (ZI) and continual learning (CL) -- two powerful cognitive abilities. Despite this importance, efforts to model ZI and CL are seriously limited in several ways. First, existing models typically lack co…

Review
pending
Role
unreviewed
Read
now
arxiv Score 11.5

Implicit Assumptions About Tasks in Continual Learning

2026-10-30 · Kathrin Korte, Christian Metzner, Akshay Bedhotiya, Viet Anh Khoa Tran, Eleni Nisioti, Sebastian Risi

Research Track A · General AI

Continual learning is the study of how a model accumulates knowledge from data received over time, rather than all at once. In practice, this is typically achieved by dividing a dataset into tasks and presenting them sequentially. Explanations of the resulting measurements take into account factors such as the similari…

Review
pending
Role
unreviewed
Read
now
arxiv Score 11.5

ResMerge: Residual Policy Learning and Merging for Continual Adaptation of Pre-trained Robot Policies

2026-10-30 · Kefei Zhu, Lingyun Xu, Yue Xin, Xiaoqi Li, Caixin Guo, Xincheng Du, Yuanpei Chen, Yaodong Yang, Hao Dong

Research Track A · General AI

Continual adaptation to new tasks is essential for robots deployed in the real world, but adapting pre-trained policies remains challenging due to parameter growth and the need for continual demonstration collection. Parameter-Efficient Continual Fine-Tuning (PECFT) isolates task-specific updates, but adapter-per-task …

Review
pending
Role
unreviewed
Read
now
arxiv Score 11.3

Flide: Accelerating Decision Inference with Exact State Forking

2026-10-30 · Bobo Li, Jiajia Song, Hao Fei, Wynne Hsu, Mong Li Lee

General AI

Decision models such as Jev have emerged as a highly promising paradigm for LLM agent systems, providing lightweight and reliable categorical reasoning for rapid System One execution. However, existing generation-oriented inference engines fail to support these workloads effectively, suffering from severe prefix recomp…

Review
pending
Role
unreviewed
Read
now
arxiv Score 10.3

ACCT: Asymmetric Cross-LLM Context Transfer

2026-10-30 · Yuzhe Liu, Haolin Tian, Tonghan Wang

General AI

In real-world deployments, large language model (LLM) systems are increasingly evolving from single-model invocation toward heterogeneous deployments in which different models are dynamically invoked for inference. When the same context is reused across models, however, each newly selected target model typically has to…

Review
pending
Role
unreviewed
Read
now
arxiv Score 10.3

Advancing Clinical Trial Outcome Prediction with Domain-Specific Large Language Models

2026-10-30 · Burak Bakkar, Hana Moalla, Mohamed Salem Jedidi, Haifa Ben Messaoud, Nicolas Lopez Carranza

General AI

Clinical trial failures are a major barrier to efficient drug development, incurring substantial costs and delaying patient access to effective therapies. Structured approaches such as the Hierarchical Interaction Network (HINT) predict trial outcomes by explicitly modelling interactions between drugs, diseases and pro…

Review
pending
Role
unreviewed
Read
now
arxiv Score 10.3

AgentShield: Fortifying Web Applications Against AI Agents and Agentic Browsers

2026-10-30 · Moiz Ali, Nafis Karim, Rownak Islam, Jyotirmay Chauhan, Dhruv Kuchhal, Jason Polakis

Research Track B · General AI

LLM-powered browser agents are transforming web automation by autonomously interacting with websites and executing complex workflows. However, the effectiveness of existing anti-bot defenses against these systems remains unclear. We present a systematic empirical evaluation of modern agentic browsers against widely dep…

Review
pending
Role
unreviewed
Read
now
huggingface Score 10.0

Venus-DefakerOne2: A Vision-Forensics Foundation Model with Broader Coverage

2026-10-27 · Chenfan Qu, Yachao Liang, Mingqi Fang, Jiangwei Xie, Zhenming Wang, Baoyue Zhang, Changjiang Jiang, Zheng Chang, Fanli Jin, Xuekang Zhu, Kaiwen Feng, Junchi Li, Weihong Zhang, Haoyue Wang, Chenfeng Zhang, Yinan Xia, Fangming Cui, Dong Fan, Jingjing Liu, Tianyi Zhang, Luobin Chen, Ji-Zhe Zhou, Sheng Li, Wei Wang, Bin Li, Ran He, Xinpeng Zhang, Linna Zhou, Jian Liu, Weiqiang Wang

Research Track A · General AI

Real-world forgery is inherently heterogeneous: it manifests not only as digital manipulation but also as physical spoofing, and is carried not only by images but also by videos. Despite substantial progress, a unified anti-forgery foundation model covering such a broad threat spectrum remains absent. In this paper, we…

Review
pending
Role
unreviewed
Read
soon
huggingface Score 9.5

GLOW-FM: Efficient Generative Low-Light Image Enhancement with One-Step Latent Flow Matching

2026-10-22 · Ariel Lapid, Yuval Haitman, Reuven Peretz, Arnon Netzer, Idit Diamant

General AI

Low-light image enhancement (LLIE) aims to recover well-exposed and perceptually faithful images from severely degraded observations. Recent diffusion-based generative methods achieve strong perceptual quality by leveraging powerful image priors, but their iterative sampling incurs substantial computational cost, limit…

Review
pending
Role
unreviewed
Read
soon
arxiv Score 9.3

CRoS: Causality Reward on Scrambled CoTs Improves Accuracy and Reasoning Faithfulness

2026-10-30 · Xiao Yu, Justin Chih-Yao Chen, Zaid Khan, Hyunji Lee, Mohit Bansal

General AI

Reinforcement learning with verifiable rewards (RLVR) has become the standard method for training language models to reason. However, RLVR typically relies solely on outcome correctness, which does not guarantee that the CoT causally contributes to the final answer. Without this connection, outcome rewards may fail to …

Review
pending
Role
unreviewed
Read
soon
arxiv Score 9.3

Movens-Zero: Scaling Humanoid Motion Foundation Models with Large-Scale Human Videos

2026-10-30 · Wendong Bu, Tinghong Ye, Qizhou Wang, Jiacheng Li, Zhiqi Ge, Yuze Lin, Zhongqi Yue, Yi Su, Jingyu Wang, Kai Shen, Jingdong Liu, Chengyuan Deng, Xiaoxuan Chen, Xin Tan, Jiaming Ji, Wenqiao Zhang, Chenghong Jiang, Tuoyu Li, Siliang Tang, Jun Xiao, Tat-Seng Chua, Yueting Zhuang, Xingxing Wang, Juncheng Li

General AI

Scaling humanoid motion foundation models is constrained by the scarcity of diverse, robot-executable data and by the difficulty of jointly learning from heterogeneous multimodal inputs while supporting real-time closed-loop control. We present Movens-5M, constructed from raw Internet videos through clip generation, qu…

Review
pending
Role
unreviewed
Read
soon
huggingface Score 8.5

VoxLock: Target-Speaker Extraction for Voice Agents Without an Enrollment Step

2026-10-29 · Arbaz Siddiqui

General AI

A voice agent's speech recognizer transcribes everyone near the microphone, including people other than the caller. I fine-tuned a public causal target-speaker extraction model (REAL-TSE, 27M parameters) on one GPU to work from a voice sample cut from the call itself, and released it as VoxLock. I score it by what a re…

Review
pending
Role
unreviewed
Read
soon
arxiv Score 8.3

Sensorium: An Open-Source Multimodal Data Collection and Curation Library for Embodied Surgical AI

2026-10-30 · Lorenzo Mazza, Ariel Rodriguez, Stefanie Speidel

General AI

Embodied surgical AI requires reliable multimodal data, yet surgical data collection remains costly, difficult to repeat, and strongly dependent on the hardware setup. We present Sensorium, an open-source Python library for recording, processing, annotating, and curating multimodal surgical data from configurable robot…

Review
pending
Role
unreviewed
Read
soon
arxiv Score 7.8

Quantum 3xplorer: AI-Native Engine for Next-Generation Quantum Materials Modeling

2026-10-30 · Vsevolod Biryukov, Sergey V. Barabash, Pranab Kumar Das, Jack Deslippe, Mauro Del Ben, Felipe H. da Jornada, Yuming Shi, Johannes Voss, Zhenglu Li, Piero Altoe, Kamal Choudhary, Ruth Pachter, Steven G. Louie, Marvin L. Cohen, Timur Bazhirov

General AI

Quantum-mechanical approaches, such as density functional theory (DFT), underpin much of modern computational materials science and are the primary source of data for AI-driven discovery. Yet the established electronic-structure codes today represent decades-old software architectures, making new physical models slow a…

Review
pending
Role
unreviewed
Read
soon
huggingface Score 7.5

Think Twice: The Metis-1.6 Technical Report

2026-10-27 · Giulianno Vollmer

General AI

Metis is a bet that better data, better training methods and new architectures can buy what a bigger parameter count cannot, at a fraction of the cost. Metis-1.6 is its latest step: two language models built from scratch by a single researcher, Praxis (2.94B parameters, 0.46B active per pass) and Logos (10.23B paramete…

Review
pending
Role
unreviewed
Read
soon
huggingface Score 7.5

rTaste: Do Research Agents Have Taste? Evaluating Early Stage Research Judgment

2026-10-29 · Art Kanke

General AI

Do research agents have taste, the judgment of which experiments are worth running before their outcomes are known? We give six frontier models seven recent NLP papers. Each model, working as a research agent alone or as a team, proposes minimum viable experiments (MVEs), forecasts whether each will pass a threshold it…

Review
pending
Role
unreviewed
Read
soon
arxiv Score 7.3

Prune Gradually, Adapt Sparsely: Efficient Pruning of Large Language Models at Extreme Sparsity

2026-10-30 · Jean Lucien Randrianantenaina, Hung Nguyen, Nasib Ullah, Jinbin Zhang, Nilesh Gupta, Inderjit S. Dhillon, Rohit Babbar

General AI

Weight pruning can reduce the storage and inference cost of large language models (LLMs), but preserving quality at high sparsity levels remains challenging. Existing methods face a quality-cost trade-off in this regime. One-shot pruning is cheap but degrades sharply as sparsity increases, even when followed by post-pr…

Review
pending
Role
unreviewed
Read
soon
arxiv Score 7.3

Same Question, Different Answers: Auditing Persona-Dependent Variation in Six LLMs on Smoking-Cessation Advice

2026-10-30 · Artin Yousefi, Aizierjiang Aiersilan, Lorien C Abroms, David A Broniatowski

General AI

Hundreds of millions of people now ask large language model (LLM) chatbots for health advice, which raises a question that is simple to pose and hard to answer: does everyone get the same advice? We audit six widely deployed LLMs from four providers on smoking cessation, a domain with codified clini- cal guidance and s…

Review
pending
Role
unreviewed
Read
soon
arxiv Score 6.3

A Multimodal Type 2 Diabetes Registry on DHIS2: A Design and Feasibility Study with the AI-READI Dataset

2026-10-30 · Saptarshi Purkayastha, Aiswarya Nandan Perumbilly, Zhenan Yin, Sherri L. Bucher

General AI

Care for type 2 diabetes increasingly draws on continuous glucose monitors, wearable devices, home sensors, clinical records and retinal images, yet few open platforms have been shown to hold all of these for the same person. We asked three questions of DHIS2, the open-source health information system that ministries o…

Review
pending
Role
unreviewed
Read
soon
arxiv Score 6.3

Dittany: Contribution-Guided Repair Framework for MAS Applications

2026-10-30 · Hanqi Sun, Yuchen Shao, Hanyi Si, Beibei Xu, Lisheng Wang, Chengcheng Wan

General AI

LLM-based multi-agent system (MAS) applications can produce useful intermediate results while leaving task requirements unsatisfied. Repairing such executions requires identifying the work still needed and the completed results that can support it. We present \tool, a contribution-guided runtime repair framework for MA…

Review
pending
Role
unreviewed
Read
soon
arxiv Score 6.3

Know When the Equation Is Wrong: Structural Uncertainty for LLM-Guided Symbolic Regression under Extrapolation

2026-10-30 · Mostafa Cham, Bayu Adhi Tama, Jianwu Wang

General AI

Large language models (LLMs) can help discover mathematical equations from data, but most LLM-guided symbolic regression methods focus on a single best equation without quantifying uncertainty over alternative functional forms. This is challenging because distinct equations can fit the observed data similarly while pro…

Review
pending
Role
unreviewed
Read
soon
huggingface Score 5.5

Supervising the Visible State: Separating Representation from View Selection in Streaming Video

2026-10-26 · Yihao Hu, Pan Wang, Yitong Lin, Zhihao Wen, Deyuan Liu, Hang Wang, Mingda Wan, Yiwen Zhao, Yutong Li, Yuchen Mao, Tao Lin

General AI

An action seen once and the same action seen three times receive the same action-recognition label, so that supervision does not require a model to count. Answers shared across histories that differ in event order, accumulation, or relations leave those distinctions unconstrained; we call this state under-specification…

Review
pending
Role
unreviewed
Read
later
arxiv Score 5.3

How BERT-based Cross-Encoders Model Relevance in Information Retrieval?

2026-10-30 · Mathias Vast, Basile Van Cooten, Laure Soulier, Benjamin Piwowarski

General AI

Cross-encoders are the most effective but also the least efficient models in Information Retrieval. Despite their success, the mechanisms underlying their relevance predictions remain poorly understood, particularly how they process to match queries and documents. This lack of interpretability limits their optimization…

Review
pending
Role
unreviewed
Read
later
arxiv Score 5.3

Latent Sparse Attention: Towards Subquadratic Ultra-Long Context Attention

2026-10-30 · Ales Kubicek, Andrea Caciolai, Liangzu Peng, Sukrit Sumant, Benjamin Hoover, Fraser Greenlee, Ruairidh M. Battleday, Radim Urban, Torsten Hoefler, James C. R. Whittington

General AI

Agentic workflows are increasingly shifting towards long-horizon tasks, such as open-ended scientific work, where models benefit from long context windows. However, at ultra-long context (above 1M tokens), full attention is prohibitively expensive. A promising approach is sparse attention, in which a dedicated indexer …

Review
pending
Role
unreviewed
Read
later
huggingface Score 4.5

No Channels Left Behind: Understanding Diffusion Acceleration through Residual Gates

2026-10-29 · Junha Park, Jaemyung Yu, Jaehui Hwang, Jong-Seok Lee, Dongyoon Han

General AI

Training acceleration methods for diffusion transformers, such as REPA and dispersive regularization, are commonly explained by additional supervision, spatial-structure priors, or improved feature geometry. We revisit this view and uncover a striking internal change: the AdaLN-Zero residual gate α, which directly cont…

Review
pending
Role
unreviewed
Read
later
arxiv Score 4.3

Counting on Dummies: Privacy Amplification with Observable Counts and Truncated Batches

2026-10-30 · Vitaly Feldman, Audra McMillan, Kunal Talwar

General AI

In many practical deployments of Differentially Private Stochastic Gradient Descent (DP-SGD), the server processes data in batches. Randomly sampling these batches allows for better privacy bounds via privacy amplification results. Most sampling schemes with provable privacy amplification, including Poisson sampling an…

Review
pending
Role
unreviewed
Read
later
arxiv Score 4.3

From Scalar Games to Unconstrained Network Zero-Sum Games: Transfer Principles for Gradient-Based Learning

2026-10-30 · Gaoqi Xue, James P. Bailey

General AI

We show that, for a broad class of gradient-based learning rules, scalar zero-sum games form a complete test family for convergence in unconstrained network zero-sum games. Specifically, suppose a learning rule achieves last-iterate or time-average convergence on every scalar game of magnitude $a$ whenever $η\leq f(a)$…

Review
pending
Role
unreviewed
Read
later
arxiv Score 4.3

Reconstruct everything everywhere all at once: An end-to-end model for future colliders

2026-10-30 · Max Hart, Kwok Yiu Wong, Samuel Van Stroud, Philippa Duckett, Tim Scanlon, Gabriel Facini

General AI

We present the Reconstruct Everything Model (REM), an end-to-end multitask model for sparse, multimodal sensor data, demonstrated in high-energy physics. To our knowledge, REM is the first demonstration of a single learned framework simultaneously reconstructing multiple particle types directly from hit-level measureme…

Review
pending
Role
unreviewed
Read
later
arxiv Score 4.3

Strategic Tanking in Stable Matching Markets

2026-10-30 · James P. Bailey

General AI

Motivated by a broad literature in psychology, organizational behavior, and matching markets, we introduce strategic tanking: an agent deliberately makes themselves less attractive to selected potential partners in order to improve their eventual assignment. Unlike traditional manipulation in stable matching, a tanker …

Review
pending
Role
unreviewed
Read
later
arxiv Score 4.3

Where ADS Crashes Differ: A Harmonized Comparison Against Human Baselines

2026-10-30 · Doyeon Jang

General AI

Crash-rate comparisons between automated driving systems (ADS) and human drivers require an exposure denominator that only a few operators publish. We instead ask a composition question: given a crash, does the mix of crash types differ between ADS and human-driven vehicles? The question needs no mileage and can theref…

Review
pending
Role
unreviewed
Read
later