Every paper checked — nothing hidden
All papers
Every paper checked for brief #4 of 12–19 Oct 2026 — nothing is hidden. Brief papers first.
All17,312In brief24More (on-topic)3,333Not relevant13,753Not assessed87Older351Could not confirm93Blogs with no paper266All blogs284
| Paper | Topic | Relevance | Feedback |
|---|---|---|---|
| CopRoute: Dependence-Aware Conformal Routing for Language Models Proposes a conformal routing method that accounts for model failure dependencies to optimize multi-model query coverage under risk budgets. Ranked #23New way to see a problemPreprint7/10 suggested | Small and efficient models via arxiv · 12 Oct, 16:29 | On-topic Designs a risk-controlled routing policy to optimize inference calls and model selection efficiency. | |
| Fly-by-Code: Embodied Coding Agents for Aerial Manipulation with Active Visual and Physical Feedback Presents an embodied coding agent framework using active physical and visual probing to refine aerial manipulation policies. Ranked #23New way to see a problemPreprint4/10 suggested | Computer vision for industrial applications via arxiv · 12 Oct, 16:28 | On-topic Focuses on robotic aerial manipulation using visual feedback in physical environments. | |
| One Word Opens the Gate: The Option-Channel Attack on Typed Decision Models as Agent Guardrails Evaluates typed decision models used as agent guardrails, uncovering vulnerabilities and asymmetric error rates in allow-or-block actions. Ranked #23Surprising resultPreprint8/10 suggested | AI agents via arxiv · 12 Oct, 16:28 | On-topic Directly assesses typed decision models serving as guardrails in multi-step agent tool-calling systems. | |
| SWE-DeceptionBench: Benchmarking Deceptive Behaviors of LLMs Across the Software Development Lifecycle Introduces SWE-DeceptionBench, a benchmark measuring active and passive deceptive model behaviors across software development lifecycle tasks. Ranked #23New testPreprint8/10 suggested | LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28 | On-topic Introduces SWE-DeceptionBench, a new benchmark evaluating deception and trustworthiness across software lifecycle stages. | |
| Learning When to Think: Utility-Guided Adaptive Reasoning for Efficient Text Retrieval Proposes an adaptive reasoning framework that supervises intermediate retrieval states and stops reasoning steps dynamically to cut latency. Ranked #24Surprising resultPreprint7/10 suggested | Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17 | On-topic Optimizes reasoning-conditioned embeddings and retrieval efficiency for search. | |
| Receiver Decides Agent-to-Agent Negotiations Studies agent-to-agent negotiations over money, finding that the receiving agent's characteristics primarily dictate the final transfer amounts. Ranked #24Surprising resultPreprint8/10 suggested | AI safety and reliability in deployment via arxiv · 15 Oct, 06:17 | On-topic Studies agent vulnerability to financial extraction and persuasion in autonomous deployments. | |
| Echo: Error-Compensated 4-Bit Tensor-Parallel Communication for LLM Training and Inference The authors develop an error-compensated 4-bit all-reduce algorithm to reduce communication overhead during distributed model training and inference. Ranked #24New way to see a problemPreprint7/10 suggested | Small and efficient models via arxiv · 15 Oct, 06:17 | On-topic Introduces 4-bit error-compensated communication to make distributed LLM inference faster. | |
| DreamTrue: Action-Faithful Robot World Model with Counterfactual Post-Training Introduces a robot world model using image-rendered actions and counterfactual post-training to improve action-following and contact prediction. Ranked #24New way to see a problemPreprint4/10 suggested | Computer vision for industrial applications via arxiv · 12 Oct, 16:28 | On-topic Builds cross-embodiment robot world models for action-conditioned manipulation perception. | |
| Auditing VLM Benchmarks from Existing Evaluation Runs Audits vision-language model benchmarks by analyzing saved evaluation artifacts and removing image inputs to detect ungrounded questions. Ranked #24New way to see a problemPreprint7/10 suggested | Multimodal models via arxiv · 12 Oct, 16:28 | On-topic Directly audits vision-language model benchmarks to identify multimodal image-text dependency failures. | |
| Safe Actions Alone Do Not Ensure Safe Agents: Identifying Unfulfilled Obligations with Guard Models Introduces a benchmark evaluating guard models on identifying unfulfilled safety obligations in language model agent trajectories alongside forbidden actions. Ranked #24New testPreprint8/10 suggested | AI agents via arxiv · 12 Oct, 16:28 | On-topic Analyzes safety mechanisms and guard models specifically for LLM-based autonomous agents. | |
| All Verdicts are Not Equal: Rethinking LLM Judge Reliability Audits the reliability of LLM judges across multiple prompt formats, temperatures, and orderings, introducing a trustworthy verdict metric. Ranked #24Surprising resultPreprint8/10 suggested | LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28 | On-topic Conducts a comprehensive reliability audit of LLM-as-a-judge across order bias, variance, and rubrics. | |
| ProgramGym: Scaling Training Environments for Long-Horizon Program Reproduction Agents Introduces ProgramGym, a framework providing verifiable sandbox environments and black-box test feedback for training long-horizon program reproduction agents. Ranked #24New way to see a problemPreprint7/10 suggested | LLMs for code via arxiv · 12 Oct, 16:28 | On-topic Directly targets software engineering agents building programs from scratch across multiple programming languages. | |
| Librarian Agents: Training LLMs to Curate Long-Context Workspaces that Anticipate Future Queries The paper trains language model agents to reorganize unstructured document collections into structured workspaces to speed up future multi-document queries. Ranked #25New way to see a problemPreprint7/10 suggested | Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17 | On-topic Proposes proactive restructuring of unstructured corpora into navigable workspaces for knowledge retrieval. | |
| Smartphone-Based, Field-Deployed AI Inspection System for Power Distribution Infrastructure Presents a field-deployed smartphone inspection system combining deep learning and geometry to assess power distribution infrastructure conditions. Ranked #25New way to see a problemPreprint4/10 suggested | Computer vision for industrial applications via arxiv · 15 Oct, 06:18 | On-topic Deploys computer vision to inspect physical infrastructure and detect field defects. | |
| Lumen-bench: Cross-Lingual Safety Gaps and the Say-Do Problem in Tool-Calling Large Language Models The paper introduces a multilingual benchmark evaluating safety and discrepancies between verbal refusals and tool actions in tool-calling language models. Ranked #25New testPreprint8/10 suggested | AI agents via arxiv · 15 Oct, 06:18 | On-topic Studies ethical violations and refusal behavior when LLMs execute actions using external tools. | |
| SpecGuardBench: Can Coding Agents Construct Reliable Validation Harnesses from Natural Language Specifications? This paper introduces a benchmark to test if coding agents can build reliable validation test harnesses from natural language specifications. Ranked #25New testPreprint8/10 suggested | AI safety and reliability in deployment via arxiv · 15 Oct, 06:17 | On-topic Highlights reliability gaps in test verification for deployed coding agents. | |
| Unconstrained CTI-REALM with Open-Weights This paper evaluates open-weight language models against closed frontier models on a cybersecurity detection benchmark using an interactive telemetry environment. Ranked #25Surprising resultPreprint7/10 suggested | Small and efficient models via arxiv · 15 Oct, 06:17 | On-topic Compares performance and inference costs of running open-weight versus closed frontier models. | |
| A Decade in a Day: Synthesizing Agent-Native Interfaces from Human-Oriented Software Evolution The paper introduces a benchmark measuring how software interface versions affect the task performance of desktop language model agents. Ranked #25New way to see a problemPreprint7/10 suggested | LLMs for code via arxiv · 15 Oct, 06:17 | On-topic Analyzes software interfaces and synthesis to support automated agents interacting with software. | |
| ConClick: Element-Wise Contrastive Learning Improves GUI Grounding Proposes a contrastive learning framework to help GUI agents distinguish target interface elements from plausible distractor elements. Ranked #25New way to see a problemPreprint7/10 suggested | Multimodal models via arxiv · 12 Oct, 16:28 | On-topic Enhances vision-language models for multimodal graphical user interface grounding. | |
| Structure Tax: How Structured Output affects LLMs Performance Evaluates how schema ordering impacts LLM performance under structured output constraints, showing reasoning-first designs prevent accuracy drops. Ranked #25Surprising resultPreprint8/10 suggested | LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28 | On-topic Evaluates model performance across schemas, benchmarking the accuracy and calibration impacts of structured output formats. | |
| SWE-at-Scale: Scalable Issue Synthesis from Historical Bug Fixes The authors propose a method to generate realistic software engineering benchmarks at scale by inverting historical bug fixes into issues. Ranked #26New testPreprint7/10 suggested | LLMs for code via arxiv · 17 Oct, 06:19 | On-topic Presents a framework to synthesize software engineering bug-fixing tasks for coding agents. | |
| ANCHOR: Assessing Naturalistic Constraint Hierarchies and Ordering in Repository-Grounded Agentic Coding This paper introduces a repository-grounded benchmark to evaluate how well coding agents follow engineering constraints and developer workflows. Ranked #26New testPreprint8/10 suggested | AI safety and reliability in deployment via arxiv · 15 Oct, 06:17 | On-topic Measures whether coding agents reliably adhere to operational and safety constraints in repositories. | |
| Scaling Inference-Time Compute with Teams of Coding Agents The paper studies compute allocation across multi-agent coding teams to identify Pareto-optimal trade-offs between cost and performance. Ranked #26Surprising resultPreprint7/10 suggested | Small and efficient models via arxiv · 15 Oct, 06:17 | On-topic Optimizes inference compute efficiency and cost reduction across heterogeneous agent teams. | |
| Trace Evolution Bench: Measuring Whether Agents Improve from Their Own Environment Traces This paper introduces a benchmark to measure whether autonomous coding agents truly improve their policies from interaction traces. Ranked #26Surprising resultPreprint8/10 suggested | AI agents via arxiv · 15 Oct, 06:17 | On-topic Examines self-evolving autonomous agents learning from execution traces in task environments. | |
| Fresh at Decision, Stale at Action: Routing New Information to LLM Decisions Already Made This paper introduces a timestamp theory and persistent ledger to route updated information to pending LLM decisions before action. Ranked #26New way to see a problemPreprint7/10 suggested | Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17 | On-topic Routes fresh evidence and background knowledge to update committed decisions. | |
| MIRAGE: Metric Inference and Reliability-Aware Gaussian Estimation from Feed-Forward Vision Anchors feed-forward visual geometry to LiDAR-inertial motion to produce metrically accurate trajectories and 3D Gaussian reconstructions. Ranked #26New testPreprint3/10 suggested | Computer vision for industrial applications via arxiv · 12 Oct, 16:28 | On-topic Combines feed-forward vision and LiDAR for industrial robot perception and 3D reconstruction. | |
| CV-OPD: Contrastive Vision On-policy Distillation Introduces Contrastive Vision On-policy Distillation to prevent vision-language models from relying on answer shortcuts rather than visual evidence. Ranked #26Surprising resultPreprint7/10 suggested | Multimodal models via arxiv · 12 Oct, 16:28 | On-topic Addresses grounding failures and token constraints specifically in vision-language models (VLMs). | |
| Verdict Without the Rule: Diagnosing and Auditing Regulatory Rule Sensitivity in LLM Compliance Systems Evaluates whether LLM compliance systems adjust verdicts when governing rules are perturbed, finding many verdicts are insensitive to rule changes. Ranked #26Surprising resultPreprint8/10 suggested | LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28 | On-topic Audits and evaluates LLM sensitivity and reliability on regulatory compliance benchmarks. | |
| Think Before You Speak: Enabling Training-Free Latent Communication Between Language Models This paper presents CacheFuse, a training-free technique enabling heterogeneous language models to communicate directly by sharing key-value cache states. Ranked #27Surprising resultPreprint7/10 suggested | Small and efficient models via arxiv · 15 Oct, 06:18 | On-topic Presents a training-free KV-cache interpolation method to bypass text communication latency. | |
| TurnCost: Turns, Not Observations: Cache-Aware Loop Engineering for Coding Agents Analyzes coding-agent trajectories to model execution costs under prompt caching, showing turn count matters far more than observation token length. Ranked #27Surprising resultPeer-reviewed8/10 suggested | AI agents via arxiv · 15 Oct, 06:17 | On-topic Analyzes iterative tool-use loops, context management, and costs in coding agents. | |
| Avoiding What Your Teacher Already Knows: Entity Familiarity for Deep-Search Distillation Measures teacher entity familiarity to filter out uninformative search trajectories when distilling deep-search capabilities into smaller agents. Ranked #27New way to see a problemPreprint7/10 suggested | Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17 | On-topic Addresses information-seeking search behavior and evidence verification across entity knowledge graphs. | |
| PreciseCoder: Faithful and Generalizable Debugging Beyond Unit-Test Rewards The paper introduces a debugging training algorithm that rewards localized semantic code edits alongside unit test execution. Ranked #27Surprising resultPreprint7/10 suggested | LLMs for code via arxiv · 15 Oct, 06:17 | On-topic Trains models for precise software debugging and repository-level program repair via localized edits. | |
| On the Risks of using LLM-Generated Tests for Regression Testing Investigates the risks of using LLM-generated regression tests, showing they can inadvertently preserve and enforce code faults. Ranked #27Surprising resultPreprint8/10 suggested | AI safety and reliability in deployment via arxiv · 12 Oct, 16:29 | On-topic Investigates the failure modes and risks of deploying LLM-generated tests that enforce faulty behavior. | |
| Golden Anchor for Video Generation Studies how text-to-image anchor properties correlate with downstream video generation quality in image-to-video diffusion pipelines. Ranked #27Surprising resultPreprint7/10 suggested | Multimodal models via arxiv · 12 Oct, 16:28 | On-topic Addresses text-to-image and image-to-video generation across visual and textual modalities. | |
| One Word Opens the Gate: The Option-Channel Attack on Typed Decision Models as Agent Guardrails Evaluates typed decision models used as agent guardrails, uncovering vulnerabilities and asymmetric error rates in allow-or-block actions. Ranked #27Surprising resultPreprint8/10 suggested | LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28 | On-topic Evaluates and benchmarks guardrail decision models across prompt injection and jailbreak tests. | |
| Beyond Actions: Learning Future Operation Targets for Vision-Language-Action Models Introduces AFCE, training vision-language-action policies to predict future physical effects rather than raw action chunks directly. Ranked #27New way to see a problemPreprint3/10 suggested | Computer vision for industrial applications via arxiv · 12 Oct, 16:28 | On-topic Develops vision-language-action policies for dexterous robotic manipulation and physical coordination. | |
| Diviner: Agentic RCA with Human-Governed Adaptation for Large-Scale Infrastructure Presents an agentic system with human-governed adaptation to automate root-cause analysis across complex, large-scale infrastructure environments. Ranked #28New way to see a problemPreprint8/10 suggested | AI agents via arxiv · 17 Oct, 06:17 | On-topic Deploys multi-agent systems for root-cause analysis in large infrastructure using tools. | |
| Plan or Retrieve First? A Scale Study of Tool Discovery Architectures Compares retrieve-then-plan and plan-then-retrieve strategies for tool discovery across varying tool registry complexities. Ranked #28Surprising resultPreprint7/10 suggested | Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06 | On-topic Studies retrieval architectures for indexing and selecting tool definitions from large registries. | |
| More Reasoning Is Not Always Better: Executing Clinical Guidelines with LLMs and Rules MedDEx combines language models for patient data extraction with deterministic rule engines to execute clinical treatment guidelines accurately and efficiently. Ranked #28Surprising resultPreprint7/10 suggested | Small and efficient models via arxiv · 15 Oct, 06:18 | On-topic Reduces inference costs and computational overhead by replacing complex reasoning with deterministic rules. | |
| Diff Diffusion: Predicting Edit Paths with Diffusion Models for Efficient Code Editing Proposes Diff Diffusion, a diffusion-based model that predicts localized code edit sequences rather than regenerating entire source files. Ranked #28New way to see a problemPreprint7/10 suggested | LLMs for code via arxiv · 15 Oct, 06:17 | On-topic Proposes a diffusion-inspired LLM framework for predicting code edit paths in software development. | |
| In-Context Learning Can Help Vision-Language Models Overcome Training Priors Demonstrates that visual in-context demonstrations help vision-language models overcome conflicting pretraining priors on counterfactual tasks. Ranked #28Surprising resultPreprint7/10 suggested | Multimodal models via arxiv · 12 Oct, 16:28 | On-topic Studies overcoming training priors in multimodal vision-language models using visual in-context learning. | |
| SimVTA: Simulation-Grounded Video-Trajectory Alignment for Driving World Action Models Proposes a simulation-grounded framework rendering real-world 3DGS scenes to supervise autonomous driving world action models. Ranked #28Preprint0/10 suggested | Computer vision for industrial applications via arxiv · 12 Oct, 16:28 | On-topic Addresses autonomous driving world action models and trajectory optimization using visual simulation. | |
| Safe Actions Alone Do Not Ensure Safe Agents: Identifying Unfulfilled Obligations with Guard Models Introduces a benchmark evaluating guard models on identifying unfulfilled safety obligations in language model agent trajectories alongside forbidden actions. Ranked #28New testPreprint8/10 suggested | LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28 | On-topic Introduces ObligationBench, a new benchmark to evaluate guard models on agent trajectories. | |
| Same Query, Different Answer: Run-to-Run Variance in Semantic Query Processing Measures output variability across repeated runs at temperature zero in semantic query engines backed by language models. Ranked #28Surprising resultPreprint8/10 suggested | AI safety and reliability in deployment via arxiv · 12 Oct, 16:28 | On-topic Documents run-to-run nondeterminism and reliability failures in production semantic query systems. | |
| Approve What, Exactly? Effect Binding and Trusted Rendering for Agent Step-Up Authorization Analyzes authorization security flaws in agent human-in-the-loop approval mechanisms and proposes a tamper-resistant approval gateway. Ranked #29Surprising resultPreprint8/10 suggested | AI agents via arxiv · 17 Oct, 06:17 | On-topic Addresses tool execution, step-up authorization, and action approval protocols in agentic systems. | |
| LLM-Based Policy Compliance Verification of Loan Applications The paper describes an automated system using language models to verify corporate loan applications against internal bank lending policies. Ranked #29New way to see a problemPreprint7/10 suggested | Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06 | On-topic Extracts compliance rules from policy documents to evaluate lending loan applications. | |
| Mind the Temporal Gap: Evaluating Video Models for Temporal Understanding in Driving Evaluates prompting methods and visual feature probing for temporal ego-motion reasoning in driving video-language models. Ranked #29Surprising resultPreprint7/10 suggested | Multimodal models via arxiv · 12 Oct, 16:28 | On-topic Evaluates multimodal video-language models on temporal dynamic understanding tasks. | |
| Dream to Remember: Continual Steering of Generalist Robot Policies with World Models Uses a frozen world model to continually steer generalist robot policies onto new tasks without catastrophic forgetting. Ranked #29Preprint0/10 suggested | Computer vision for industrial applications via arxiv · 12 Oct, 16:28 | On-topic Studies continual learning and world models for generalist robot policy execution. | |
| AI Models Can Provably Hide Arbitrary Capabilities The paper shows that backdoor mechanisms can cryptographically hide arbitrary neural network capabilities until activated by secret triggers. Ranked #29Surprising resultPreprint8/10 suggested | LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28 | On-topic Reveals fundamental limits and failure modes of elicitation-based capability audits and evaluations. | |
| Beyond Binary Verifiable Rewards: Behavioral Self-Rewarding for SWE Agents Introduces behavioral self-rewarding to refine reinforcement learning feedback for software engineering agents by comparing execution trajectories. Ranked #29New way to see a problemPreprint7/10 suggested | LLMs for code via arxiv · 12 Oct, 16:28 | On-topic Focuses on training software engineering (SWE) agents to improve debugging and coding trajectories. | |
| Defence against the dark arts of distillation: Unsupervised discovery of subliminal learning Introduces an unsupervised method to identify and remove subliminally transferred traits from model parameter updates during knowledge distillation. Ranked #29Surprising resultPreprint7/10 suggested | Small and efficient models via arxiv · 12 Oct, 16:28 | On-topic Addresses model distillation mechanics and trait inheritance when transferring capabilities between models. | |
| SWE-DeceptionBench: Benchmarking Deceptive Behaviors of LLMs Across the Software Development Lifecycle Introduces SWE-DeceptionBench, a benchmark measuring active and passive deceptive model behaviors across software development lifecycle tasks. Ranked #29New testPreprint8/10 suggested | AI safety and reliability in deployment via arxiv · 12 Oct, 16:28 | On-topic Directly assesses deceptive behaviors, honesty risks, and reliability of LLMs under external inducements. | |
| Which Knowledge Should LLMs Trust? Uncovering a Generalizable Steering Direction Across Knowledge Conflicts The paper identifies an activation steering vector that controls whether language models rely on internal memory or retrieved external context. Ranked #30Surprising resultPreprint7/10 suggested | Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06 | On-topic Directly investigates steering models to trust retrieved external context over memory. | |
| Lumen-bench: Cross-Lingual Safety Gaps and the Say-Do Problem in Tool-Calling Large Language Models The paper introduces a multilingual benchmark evaluating safety and discrepancies between verbal refusals and tool actions in tool-calling language models. Ranked #30New testPreprint8/10 suggested | LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18 | On-topic Introduces Lumen-bench, a multi-lingual benchmark evaluating tool-calling refusal behavior across 30 models. | |
| Knowing the User Is Not Knowing What We Established: Grounding Awareness in Personal Agents The authors introduce a benchmark to evaluate whether personal agents can distinguish accepted user arrangements from unacknowledged proposals. Ranked #30Surprising resultPreprint8/10 suggested | AI agents via arxiv · 15 Oct, 06:17 | On-topic Evaluates and tracks whether personal conversational agents properly ground user authorization before taking actions. | |
| TestPrism: Rethinking Test Evaluation Beyond a Single Reference Proposes evaluating generated test suites across pools of valid and invalid implementations rather than against a single reference implementation. Ranked #30New testPreprint7/10 suggested | LLMs for code via arxiv · 12 Oct, 16:29 | On-topic Studies LLM coding agents generating test suites and benchmarks multi-reference test evaluation. | |
| What 30,000 Hours of Ego-centric Video Does Not Teach Studies how scaling to 30,000 hours of egocentric video affects world model fidelity for agents versus object interactions. Ranked #30Surprising resultPreprint7/10 suggested | Multimodal models via arxiv · 12 Oct, 16:28 | On-topic Evaluates multimodal world models trained on large-scale ego-centric video for agent-object interaction. | |
| RARA: Robust Policy Fine-Tuning through Retrieval-Anchored Representation Alignment for Robotic Manipulation Aligns robot policy representations with retrieved pretraining examples to maintain generalization during low-data fine-tuning. Ranked #30Preprint0/10 suggested | Computer vision for industrial applications via arxiv · 12 Oct, 16:28 | On-topic Focuses on robotic manipulation policy fine-tuning, directly relevant to industrial robotics perception and action. | |
| Unsupervised Multimodal Reward Modeling Trains an unsupervised vision-language reward model using intermediate layer representations and caption perturbations without human preference labels. Ranked #30Surprising resultPreprint7/10 suggested | Small and efficient models via arxiv · 12 Oct, 16:28 | On-topic Trains a lightweight linear head reward model at a fraction of fine-tuning compute. | |
| Reward Hacking as a Transmissible Strategy in Multi-Agent Coding Systems Demonstrates that reward-hacking strategies can propagate across multiple agents through shared memory and code in multi-agent coding systems. Ranked #30Surprising resultPreprint8/10 suggested | AI safety and reliability in deployment via arxiv · 12 Oct, 16:28 | On-topic Analyzes reward hacking and alignment vulnerabilities that propagate across interacting agents. |
181–240 of 17,312← PreviousNext →