Sens.aiAI signal desk
TodayRadarBriefing
Admin sign in
Sens.aiSensif.aiAI signal desk
ModelsResearchCountriesOpen vs ClosedComputeBetaCommentary
Loading…
TodayRadarBriefingAdmin

Radar

Research Radar

Which research topics are accelerating, and where earlier papers have already landed in shipped models.

Filtered to Agents — 75 papersClear filter

More papers match this topic — see “Show more” below.

Papers this week

50

Trending topic

Agents

▲ 19%

papers this window vs prior window

Papers linked to models

128

linked by the desk, past 90 days

Median days paper → model

—

lower is faster

Topic velocity

papers this window by topic · Δ vs prior window · a paper can carry more than one topic

AgentsAgents
134▲19%
Evaluation & BenchmarksEvaluation & Benchmarks
39▲44%
MultimodalMultimodal
23▼30%
Safety & Alignment

Topic momentum

prior window → this window

Agents

134 this window ▲19%

Evaluation & Benchmarks

39 this window ▲44%

Multimodal

23 this window ▼30%

Where research lands: paper → model linkage

Links are AI-inferred by the desk's LLM from tech reports, system cards and citations — each carries a confidence level and is not a claim by the authors.

Let Credit Follow Computation: Architecture-Aware Credit Transport for Large Language Model Reinforcement Learning

arXiv:2608.21501AI-inferred · high

KVBoost: Chunk-Level Key-Value Cache Reuse with Deviation-Guided Recomputation for Efficient Large Language Model Inference

arXiv:2608.21362AI-inferred · high

K-Bench: measuring model performance on real scientific agent requests

arXiv:2608.21601AI-inferred · high

IB-RL: Isolated Bilateral Reinforcement Learning for Strategic Dialogue Agents

Filtered to Agents — 75 papersClear filter

More papers match this topic — see “Show more” below.

Papers on Agents

Clear filter
agentic AImulti-drone systemspublished Aug 25, 2026 · arXiv:2608.21444

Agentic AI for Safety-critical Multi-drone Systems: Challenges and Opportunities

scientific artificial intelligenceagent evaluationpublished Aug 25, 2026 · arXiv:2608.21601

K-Bench: measuring model performance on real scientific agent requests

Safety & Alignment
22▲47%
Training & OptimizationTraining & Optimization
21▲50%
Efficiency & InferenceEfficiency & Inference
19▼10%
ReasoningReasoning
15▲25%
InterpretabilityInterpretability
12▼20%
Reinforcement LearningReinforcement Learning
10▼29%
Retrieval & RAGRetrieval & RAG
9▼18%
Code GenerationCode Generation
8▼27%
Privacy & SecurityPrivacy & Security
7▲250%
Language UnderstandingLanguage Understanding
5▲25%
Generative ModelsGenerative Models
5▼67%
Robotics & Embodied AIRobotics & Embodied AI
5▼29%
Long ContextLong Context
4▼20%
Speech & AudioSpeech & Audio
4▼33%
Mixture of ExpertsMixture of Experts
3▲50%
Computer VisionComputer Vision
3▼63%
Data & Synthetic DataData & Synthetic Data
1▼75%
OtherOther
165▲12%
View as table
TopicPapers this windowPrior windowΔ
Agents134113+19%
Evaluation & Benchmarks3927+44%
Multimodal2333-30%
Safety & Alignment2215+47%
Training & Optimization2114+50%
Efficiency & Inference1921-9%
Reasoning1512+25%
Interpretability1215-20%
Reinforcement Learning1014-29%
Retrieval & RAG911-18%
Code Generation811-27%
Privacy & Security72+250%
Language Understanding54+25%
Generative Models515-67%
Robotics & Embodied AI57-29%
Long Context45-20%
Speech & Audio46-33%
Mixture of Experts32+50%
Computer Vision38-62%
Data & Synthetic Data14-75%
Other165148+12%

Safety & Alignment

22 this window ▲47%

Training & Optimization

21 this window ▲50%

Efficiency & Inference

19 this window ▼10%
arXiv:2608.06735AI-inferred · high

NiyamAI - An Intent-Bound AI Agent with Cryptographically Verifiable Guardrails using Zero-Knowledge Proofs

arXiv:2608.07167AI-inferred · high

WebGrader: Training LLMs for Web Development with Self-Evolving Programmatic Grader

arXiv:2608.06474AI-inferred · high

Divergent Response Modes in Frontier Language Models Under Steering Pressure

arXiv:2608.06578AI-inferred · high

Agent Memory Distillation: Empowering Small LLM Agents with Hierarchical Teacher Memory

arXiv:2608.07169AI-inferred · high

Reason Wide, Not Deep: Amortizing the Reasoning Premium into Distilled Skills

arXiv:2608.07885AI-inferred · high

SodaMem: Evidence-Grounded Temporal Graph Memory for LLM Agents

arXiv:2608.08055AI-inferred · high

Who Verifies the Benchmark? Decentralizing Trust in Large Language Model Evaluation

arXiv:2608.07762AI-inferred · high
Qwen 3.8 Max2 links
claude-fable-5, claude-sonnet-5, claude-opus-52 links
GPT-5.64 links
DeepSeek-V4-Flash-07313 links
GPT-5.6 Luna2 links
GLM-5.31 link
View as table
PaperLinked modelRelationConfidence
Let Credit Follow Computation: Architecture-Aware Credit Transport for Large Language Model Reinforcement Learning (2608.21501)Qwen 3.8 Maxevaluates99%
KVBoost: Chunk-Level Key-Value Cache Reuse with Deviation-Guided Recomputation for Efficient Large Language Model Inference (2608.21362)Qwen 3.8 Maxevaluates98%
K-Bench: measuring model performance on real scientific agent requests (2608.21601)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates100%
K-Bench: measuring model performance on real scientific agent requests (2608.21601)GPT-5.6evaluates100%
IB-RL: Isolated Bilateral Reinforcement Learning for Strategic Dialogue Agents (2608.06735)DeepSeek-V4-Flash-0731evaluates95%
NiyamAI - An Intent-Bound AI Agent with Cryptographically Verifiable Guardrails using Zero-Knowledge Proofs (2608.07167)GPT-5.6evaluates98%
WebGrader: Training LLMs for Web Development with Self-Evolving Programmatic Grader (2608.06474)GPT-5.6evaluates99%
Divergent Response Modes in Frontier Language Models Under Steering Pressure (2608.06578)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates99%
Divergent Response Modes in Frontier Language Models Under Steering Pressure (2608.06578)GPT-5.6evaluates99%
Agent Memory Distillation: Empowering Small LLM Agents with Hierarchical Teacher Memory (2608.07169)GPT-5.6 Lunatechnique used98%
Reason Wide, Not Deep: Amortizing the Reasoning Premium into Distilled Skills (2608.07885)GPT-5.6 Lunaevaluates98%
SodaMem: Evidence-Grounded Temporal Graph Memory for LLM Agents (2608.08055)DeepSeek-V4-Flash-0731technique used98%
Who Verifies the Benchmark? Decentralizing Trust in Large Language Model Evaluation (2608.07762)DeepSeek-V4-Flash-0731evaluates98%
Who Verifies the Benchmark? Decentralizing Trust in Large Language Model Evaluation (2608.07762)GLM-5.3evaluates98%
Who Verifies the Benchmark? Decentralizing Trust in Large Language Model Evaluation (2608.07762)GPT-5.6 Lunaevaluates98%
An Agentic AI Framework Overcomes Fundamental Limitations of Large Language Models for Glaucoma Detection from Fundus Photography (2608.07651)GPT-5.6 Lunaevaluates99%
An Agentic AI Framework Overcomes Fundamental Limitations of Large Language Models for Glaucoma Detection from Fundus Photography (2608.07651)Gemini 3.7 Flashevaluates99%
When the Judge Should Not Decide: Evidence-Locked, Non-Compensatory Selection Bounds LLM-Judge Failure in Reasoning Pipelines (2608.07813)DeepSeek-V4-Flash-0731evaluates95%
The Authority Expectancy Effect in Multi-User Conflict (2608.08026)Grok 4.6evaluates95%
The Authority Expectancy Effect in Multi-User Conflict (2608.08026)GPT-5.6 Lunaevaluates95%
The Authority Expectancy Effect in Multi-User Conflict (2608.08026)Gemini 3.7 Flashevaluates95%
The Authority Expectancy Effect in Multi-User Conflict (2608.08026)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates95%
Can Legal AI Know When It Is Wrong? And Do Students Know When It Is? (2608.21089)GPT-5.6 Lunaevaluates99%
Calibrating Criterion Revision in LLM Agents: Failure Modes and a Trace-Anchored Protocol (2608.20729)Qwen 3.8 Maxevaluates98%
Why2Speak: Faithful Reasoning for Abstaining Action Policies (2608.20670)Qwen 3.8 Maxevaluates98%
No Judgment Without a Reason: Counterfactual Receipts for Versioned AI Evaluators (2608.20938)Qwen 3.8 Maxevaluates99%
Automated Trajectory Evaluation for Mobile Agents via Step-Level Consequence Reasoning and Aggregation (2608.20797)Qwen 3.8 Maxtechnique used98%
Beyond Endpoint Gains: A Weight-Delta Audit of Medical Specialization (2608.20768)GPT-5.6 Lunaevaluates99%
Beyond Endpoint Gains: A Weight-Delta Audit of Medical Specialization (2608.20768)Qwen 3.8 Maxevaluates99%
Beyond Endpoint Gains: A Weight-Delta Audit of Medical Specialization (2608.20768)Gemma 4evaluates99%
Natural-Language-Guided Generator-Agnostic Shortlisting for Protein Binder Design (2608.20755)GPT-5.6 Lunatechnique used95%
DirEAG: Dirichlet Evidence Aggregation for Calibrating Verbalized Confidence in Mathematical Reasoning (2608.20717)Gemma 4evaluates95%
DirEAG: Dirichlet Evidence Aggregation for Calibrating Verbalized Confidence in Mathematical Reasoning (2608.20717)Mistral OCR 4evaluates95%
DirEAG: Dirichlet Evidence Aggregation for Calibrating Verbalized Confidence in Mathematical Reasoning (2608.20717)Qwen 3.8 Maxevaluates95%
Nexus: Depth-Adaptive KV-Cache Splicing and Retrieval-Decoupled Tool Routing for Agentic LLMs on Unified Memory (2608.20397)Qwen 3.8 Maxevaluates99%
Truth Lies Deep: Countering Semantic Camouflage via Latent Intent Verification (2608.20378)Gemma 4evaluates98%
Truth Lies Deep: Countering Semantic Camouflage via Latent Intent Verification (2608.20378)Qwen 3.8 Maxevaluates98%
PrimeAgentOrchestrator: Memory-Primed Agent Spawning for Personal AI Infrastructure (2608.20342)claude-fable-5, claude-sonnet-5, claude-opus-5technique used98%
Personalized Privacy Control in LLMs via Attention Head Intervention (2608.21209)Qwen 3.8 Maxevaluates99%
TreeWY: Speculative Verification for Gated DeltaNet Hybrids (2608.20961)Qwen 3.8 Maxevaluates98%
Weighted Memory Tree: Remembering What Matters for Long-Horizon LLM Agents (2608.20631)Gemma 4evaluates99%
Weighted Memory Tree: Remembering What Matters for Long-Horizon LLM Agents (2608.20631)Qwen 3.8 Maxevaluates99%
FlavourBench: Ranking Frontier Language Models with Executable Culinary Ground Truth (2608.20574)Grok 4.6evaluates99%
StateSight: Benchmarking Latent Spatial-State Reconstruction in Vision-Language Models (2608.20414)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates99%
StateSight: Benchmarking Latent Spatial-State Reconstruction in Vision-Language Models (2608.20414)GPT-5.6 Lunaevaluates99%
SkillLens: Visual Skill Cards for Retrieval-Augmented GUI Action Prediction and On-Policy Distillation (2608.10775)GPT-5.6 Lunaevaluates98%
CHORUS: Complementary Experts for High-Coverage Testbench Stimulus Generation (2608.10090)DeepSeek-V4-Flash-0731evaluates98%
DSAgentBench: Can Agents Automate End-to-End Data-Science Workflows in Real Computer Environments? (2608.10366)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates98%
Apodex Discovery: Reality Benchmarks and Environments for Evaluating and Building Discoverative Artificial Intelligence (2608.11341)GPT-5.6 Lunaevaluates98%
AutoWorldModel-Bench: A State-Centric Benchmark for Automated World-Model Research (2608.11216)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates98%
MBA: Multimodal Benchmark and Agents for Real-World Business Ideation (2608.11616)GPT-5.6 Lunatechnique used98%
When Self-Consistency Backfires: Majority Vote Hurts the Majority of Hard Science Problems for Small LLMs (2608.11403)Qwen 3.8 27Bevaluates99%
XBridge: Entity-Grounded Latent Bridge for Heterogeneous LLM Communication (2608.11676)Qwen 3.8 27Bevaluates95%
Graph-Structured Rubrics: Compiling Rubrics into Typed Evaluation Graphs for LLM Judges (2608.12097)GPT-5.6 Lunaevaluates95%
HUGIN: Enhancing Vision-Language Planning for Autonomous Logistics Sorting (2608.11692)Qwen 3.8 27Bevaluates99%
Claim-Level Reliability Assessment for Efficient Test-Time Reasoning (2608.11994)GPT-5.6 Lunaevaluates95%
From Numbers to Judgment: Specialist LLM Agents and Reinforcement Learning for European Listed Real Estate (2608.11381)Qwen 3.8 27Btechnique used98%
Retrofitting Recurrent Depth into a Pretrained Language Model: Installation, Extrapolation, Transfer, and Retention at Two Parameter Budgets (2608.11233)Qwen 3.8 27Btechnique used99%
Backtrader-Bench: Benchmarking LLM Agents on Algorithmic Trading with Self-Generated MCQs (2608.11232)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates95%
Backtrader-Bench: Benchmarking LLM Agents on Algorithmic Trading with Self-Generated MCQs (2608.11232)GPT-5.6 Lunaevaluates95%
FrontierFinance: A Challenging Benchmark for Measuring Frontier Intelligence of Finance Agents (2608.11683)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates98%
Mechanist: AI as a Scientific Instrument for Discovering the Mechanisms of Intelligence (2608.12036)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates86%
Harnessing agent memory to build lifelong AI partners for materials scientists (2608.11224)GPT-5.6 Lunaevaluates95%
Can Frontier LLMs Match Natively Multimodal Embeddings? A Comparison on Hard-Negative Text-to-Image Retrieval (2608.11343)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates99%
Can Frontier LLMs Match Natively Multimodal Embeddings? A Comparison on Hard-Negative Text-to-Image Retrieval (2608.11343)GPT-5.6 Lunaevaluates99%
An Agentic Workflow for Legacy HPC Modernization: Converting the Two-Electron-Integral Core of GAMESS (2608.12249)claude-fable-5, claude-sonnet-5, claude-opus-5technique used95%
Distribird: Literature-Informed Prior Distribution Design for Bayesian Model Calibration (2608.11210)Qwen 3.8 27Bevaluates95%
From Monolithic to Modular: Segment-level Automatic Prompt Optimization (2608.11219)GPT-5.6 Lunaevaluates98%
RecSys Factory: Bounding LLM Agent Autonomy to Decision Points in the Industrial Recommender Lifecycle (2608.11241)claude-fable-5, claude-sonnet-5, claude-opus-5technique used90%
Reasoning Jury: Multi-Model Consensus for Evaluating Reasoning Traces (2608.12585)GPT-5.6 Lunaevaluates95%
Enhancing Virtual Agents through SLMs and Edge-Computing: An Exploratory Evaluation of Think and Memory Processes (2608.13420)Qwen 3.8 27Bevaluates98%
Rethinking Normalization Placement for LLMs: Post-Norm under Curriculum Depth Growing (2608.13156)Qwen 3.8 27Bevaluates95%
Jointly Predicting Courses and Grades Using a Transformer-Based Model (2608.13409)trace-supervised symbolic neural CPUtechnique used99%
Explanatory Engagement Under Rare Anomalous Failure: Asymptotic Rarity in Model Behavior (or: The Asymptotic AI) (2608.13063)Qwen 3.8 27Bevaluates99%
ε-MemEvo: Adaptive Cross-Task Memory Transfer for LLM Program Evolution (2608.12522)AlphaEvolverelated94%
ε-MemEvo: Adaptive Cross-Task Memory Transfer for LLM Program Evolution (2608.12522)GPT-5.6 Lunaevaluates98%
Governed Persistent Memory: Source-Bound State Semantics and Fail-Closed Release for Long-Horizon Agents (2608.12476)Qwen 3.8 27Bevaluates99%
Privacy-Preserving RAG by Concealing Sensitive Information from External LLMs (2608.12675)Qwen 3.8 27Bevaluates95%
Polish Medical Visual Question Answering: Vision-Language Models Underutilize Visual Evidence (2608.12928)GPT-5.6 Lunaevaluates99%
From Passive Delegates to Strategic Negotiators: Reinforcing Social Reasoning in Small Language Models with SocialRL (2608.13787)GPT-5.6 Lunaevaluates95%
Explanation Multiplicity: Circuit-Level Interpretability Evidence Does Not Survive Defensible Analytic Variation (2608.13754)GPT-5.6 Lunaevaluates99%
Wrong but Useful: Trajectory Value Beyond Answer Correctness in Multi-Agent Messages (2608.14375)GPT-5.6 Lunaevaluates95%
MemoryLake on MemoryArena: A Matched Study of Agent Memory Backends (2608.13883)GPT-5.6 Lunaevaluates95%
Implementing Computational Law in Wolfram Language for the Governance of Artificial Intelligence (2608.13958)GPT-5.6 Lunaevaluates100%
Intern-S2-Mobius: Foundation Model with Decoupled Knowledge and Reasoning (2608.14290)Qwen 3.8 Maxrelated95%
StateM: Reaching 95.3% Raw Accuracy, or a $15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling (2608.15089)DeepSeek-V4-Flash-0731evaluates99%
StateM: Reaching 95.3% Raw Accuracy, or a $15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling (2608.15089)GPT-5.6 Lunaevaluates95%
Large Language Models Show Metacognitive Sensitivity in Medical Reasoning (2608.14552)GPT-5.6 Lunaevaluates99%
When Agentic Executions Fail: Detecting and Localizing Runtime Faults from Telemetry (2608.14680)DeepSeek-V4-Flash-0731evaluates98%
SKILL: Self-correcting Knowledge-guided Iterative Large Language Model Agent for Logic Optimization (2608.14579)GPT-5.6 Lunatechnique used98%
ACTS-SQL: Agentic and Critic-Oriented Tree-Structured SQL Correctness with Large Language Models (2608.15145)GPT-5.6 Lunaevaluates95%
The Unwritten Benchmark: A New Challenge for Multimodal Machine Learning in Abstract Perceptual Reasoning (2608.14558)GPT-5.6 Lunaevaluates98%
The Hallucination Snowball: Modeling Error Propagation as State Transitions in Multi-Agent LLM Pipelines (2608.14588)GPT-5.6 Lunaevaluates98%
LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning Tasks (2608.14927)GPT-5.6 Lunaevaluates95%
Mechanistic Tomography: Designed Measurement for Control-Oriented Interpretability (2608.19338)Qwen 3.8 Maxevaluates98%
Mechanistic Tomography: Designed Measurement for Control-Oriented Interpretability (2608.19338)GPT-5.6 Lunaevaluates98%
When AI Writes, Who Gets Cited? Evidence of Citation Monoculture Across Language Models (2608.19230)GPT-5.6 Lunaevaluates98%
MidTool: Mid-training Data Synthesis for Agentic Tool Use (2608.20314)Qwen 3.8 Maxtechnique used98%
Automatic bioinformatic software named entity recognition from literature (2608.19201)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates90%
Automatic bioinformatic software named entity recognition from literature (2608.19201)Grokevaluates90%
Automatic bioinformatic software named entity recognition from literature (2608.19201)Gemini 3.7 Flashevaluates90%
Automatic bioinformatic software named entity recognition from literature (2608.19201)GPT-5.6 Lunaevaluates90%
Can Conversational AI loosen Us-Versus-Them Boundaries? The Effects of Common, Dual, and Separate Identity Framings on Pro-Immigrant Intergroup Helping (2608.19220)GPT-5.6 Lunatechnique used99%
A Virtual Member of a Community of Practice for the Society of Petroleum Engineers: From Prototype to Deployment (2608.19199)Athena-Brain-8Bevaluates98%
Phantom Gains: Auditing Self-Improvement Against a Measured Null (2608.20290)Qwen 3.8 Maxevaluates98%
ReguSim: Evaluating LLM Agent Rule Grounding in Financial Compliance (2608.19974)Gemini 3.7 Flashevaluates95%
ReguSim: Evaluating LLM Agent Rule Grounding in Financial Compliance (2608.19974)DeepSeek-V4-Flash-0731evaluates95%
PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agents (2608.19861)Gemini 3.7 Flashevaluates99%
PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agents (2608.19861)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates99%
PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agents (2608.19861)GPT-5.6 Lunaevaluates99%
SAPO: Single-Rollout Autoregressive Policy Optimization for Agentic Reinforcement Learning (2608.19842)Qwen 3.8 Maxevaluates98%
Can Agent Memory Systems Track Evolving State? (2608.19652)Qwen 3.8 Maxevaluates98%
Can Agent Memory Systems Track Evolving State? (2608.19652)DeepSeek-V4-Flash-0731evaluates98%
From Retrieved Context to Runtime Control: Adaptive Compression for Edge-based RAG (2608.19535)Qwen 3.8 Maxevaluates99%
Air Traffic Control Using Large Language Models: Prompt Engineering, Architecture, and Evaluation (2608.19299)GPT-5.6 Lunaevaluates80%
When Personalization Becomes Bias: Structural and Discursive Religious Framing in AI-Generated Financial Advice (2608.16909)Grokevaluates98%
Different Facets of Verbalised Overconfidence: an Interpretability Study (2608.18106)Qwen 3.8 Maxevaluates99%
DeepTCM1.0: A Multi-Expert AI Agent for Deciphering Mechanisms of Chinese Herbal Formulae Based on General Large Language Models (2608.18103)DeepSeek V4-Flashtechnique used99%
Abliteration Mitigation via Refusal Aliases (2608.18093)Gemma 4evaluates99%
Nine Emotion Centroids: A Label-Free Valence Axis That Transfers Across Four Modalities (2608.18090)Gemma 4evaluates85%
Nine Emotion Centroids: A Label-Free Valence Axis That Transfers Across Four Modalities (2608.18090)Qwen 3.8 Maxevaluates85%
Nine Emotion Centroids: A Label-Free Valence Axis That Transfers Across Four Modalities (2608.18090)Mistral OCR 4evaluates85%
Latent Space Refusal Anchoring for Low-Resource African Languages: Mechanistic Safety Recovery Without Retraining (2608.18089)Qwen 3.8 Maxevaluates99%
Latent Space Refusal Anchoring for Low-Resource African Languages: Mechanistic Safety Recovery Without Retraining (2608.18089)Mistral OCR 4evaluates99%
Beyond the Transcript: Detecting Covert Co ordination in Latent Multi-Agent Communication (2608.19161)Qwen 3.8 Maxevaluates98%
Can a Lightweight Multimodal Model Estimate LLM Reasoning Performance? A Study for Compute-Optimal Document Inference (2608.18591)Qwen 3.8 Maxtechnique used100%
A Jagged Frontier: Evaluating Robustness of Code Agents to Semantics-Preserving Transformations (2608.18389)Qwen 3.8 Maxevaluates98%
A Jagged Frontier: Evaluating Robustness of Code Agents to Semantics-Preserving Transformations (2608.18389)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates98%
Governance Records as Supervision: Verifier-Selected Self-Training for Structured Workflow Repair (2608.18324)Qwen 3.8 Maxevaluates98%
Efficient Adaptation of LLMs for Hate Speech Detection in Low-Resource Languages: A Comparative Study on Roman Urdu (2608.18142)Mistral OCR 4evaluates99%
Position: Collusion Risks Among AI Reasoning Agents Justify Certification Requirements for Making Market Decisions (2608.18078)DeepSeek V4-Flashevaluates95%
Computational Orientalism: Measuring Structural Discourse Bias in Large Language Models Using the Middle East Cultural Sensitivity Score (MECSS) (2608.18100)GPT-5.6 Lunaevaluates99%
Breaking the weakest link to evade vision language models (2608.18938)Qwen 3.8 Maxevaluates99%
Training-Free Inference-Time Self-Reflection and Cost-Bounded Early Stopping for Large Language Models (2608.18884)Qwen 3.8 Maxevaluates98%
CTIFoundry: An Agent-Native Corpus Scaffold for Cyber Threat Intelligence (2608.18613)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates80%
FM-Bench: A Benchmark for Long-Horizon Management with Competing Agents (2608.18423)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates100%
Measuring the Partial-Credit Gap: A Strict Benchmark on Vietnam's 2025 Convex Marking Scheme (2608.18336)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates90%
Measuring the Partial-Credit Gap: A Strict Benchmark on Vietnam's 2025 Convex Marking Scheme (2608.18336)Qwen 3.8 Maxevaluates98%
Solving Is Not Drawing: A Benchmark for Diagrammatic Reasoning in Olympiad Geometry (2608.18111)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates80%
Solving Is Not Drawing: A Benchmark for Diagrammatic Reasoning in Olympiad Geometry (2608.18111)GPT-5.6 Lunaevaluates80%
ComponentBench: Diagnosing Component-Level Failures in Computer-Use Agents (2608.18307)Qwen 3.8 Maxevaluates99%
ComponentBench: Diagnosing Component-Level Failures in Computer-Use Agents (2608.18307)Gemini 3.7 Flashevaluates99%
ComponentBench: Diagnosing Component-Level Failures in Computer-Use Agents (2608.18307)GPT-5.6 Lunaevaluates99%
Potential of ChatGPT in predicting stock market trends based on Twitter Sentiment Analysis (2311.06273)GPT-5.6 Lunaevaluates95%
StagedWorkspace: A Versioned Workspace for Knowledge-Work Agents (2608.18050)GPT-5.6 Lunaevaluates98%
StagedWorkspace: A Versioned Workspace for Knowledge-Work Agents (2608.18050)Gemini 3.7 Flashevaluates98%
Auditing Self-Evolution in Financial Agents: Capability Gains, Security Drift, and Execution-Interface Mismatch (2608.17684)Qwen 3.8 Maxevaluates98%
The Price of Thinking: Reasoning Effort as a Model-Specific API Contract (2608.16956)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates95%
Beyond the Trace: Coupling an Interpretable Reasoning-State Readout to Native MoE Routing (2608.17638)GPT-5.6 Lunaevaluates95%
TRUSS: Towards Task-Reliable and User-Safe Automated Agent Skill Generation (2608.17588)GPT-5.6 Lunaevaluates95%
Agent Lightning v1.0: Towards Harnessed Agentic RL (2608.17528)Qwen 3.8 Maxevaluates99%
LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents (2608.17393)Qwen 3.8 Maxevaluates99%
TileMix: Tile-Centric Mixed-Precision Attention for LLM Inference Acceleration (2608.17336)Qwen 3.8 Maxevaluates98%
SignalReasoner: Assessing the Upper Bound of 3B Models for Signal Mathematical Reasoning (2608.17301)Qwen 3.8 Maxevaluates98%
When Personalization Becomes Bias: Structural and Discursive Religious Framing in AI-Generated Financial Advice (2608.16909)Grok 4.6evaluates99%
When Personalization Becomes Bias: Structural and Discursive Religious Framing in AI-Generated Financial Advice (2608.16909)Gemini 3.7 Flashevaluates98%
When Personalization Becomes Bias: Structural and Discursive Religious Framing in AI-Generated Financial Advice (2608.16909)GPT-5.6 Lunaevaluates98%
FedPref: Federated Preference Learning for Structured Radiology Report Extraction (2608.16971)Qwen 3.8 Maxtechnique used99%
Explicit State Elicitation Is Not Enough: A Controlled Audit of Memory-Policy Classification (2608.17247)GPT-5.6 Lunaevaluates98%
GxP-Agent: Process-DAG Topology for Reliable Clinical Trial Programming with LLM Agents (2608.16890)GPT-5.6 Lunaevaluates99%
GxP-Agent: Process-DAG Topology for Reliable Clinical Trial Programming with LLM Agents (2608.16890)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates99%
StateM: Reaching 95.3% Raw Accuracy, or a $15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling (2608.15089)DeepSeek V4-Flashevaluates99%
StateM: Reaching 95.3% Raw Accuracy, or a $15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling (2608.15089)GPT-5.6 Solevaluates99%
Does a Tool Result Carry More Authority Than Plain Text? Three Prospective Studies of False-Claim Adoption in a Synthetic Assignment Task with Claude Opus 5 (2608.14992)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates100%
Small Models Scout Bottleneck Order for Large-Model Data Control (2608.14936)Qwen 3.8 Maxevaluates95%
LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning Tasks (2608.14927)GPT-5.6 Soltechnique used98%
The Hallucination Snowball: Modeling Error Propagation as State Transitions in Multi-Agent LLM Pipelines (2608.14588)Qwen 3.8 Maxevaluates98%
The Hallucination Snowball: Modeling Error Propagation as State Transitions in Multi-Agent LLM Pipelines (2608.14588)GPT-5.6 Solevaluates99%
SKILL: Self-correcting Knowledge-guided Iterative Large Language Model Agent for Logic Optimization (2608.14579)Gemini 3.7 Flashtechnique used98%
SKILL: Self-correcting Knowledge-guided Iterative Large Language Model Agent for Logic Optimization (2608.14579)claude-fable-5, claude-sonnet-5, claude-opus-5technique used98%
SKILL: Self-correcting Knowledge-guided Iterative Large Language Model Agent for Logic Optimization (2608.14579)GPT-5.6 Soltechnique used99%
Anatomy of a Quantized Agent: VRAM Stability and Forecasting in Code-Synthesis Agentic Workloads (2608.15117)Qwen 3.8 Maxevaluates99%
When Agentic Executions Fail: Detecting and Localizing Runtime Faults from Telemetry (2608.14680)DeepSeek V4-Flashevaluates99%
ACTS-SQL: Agentic and Critic-Oriented Tree-Structured SQL Correctness with Large Language Models (2608.15145)GPT-5.6 Solevaluates95%
ACTS-SQL: Agentic and Critic-Oriented Tree-Structured SQL Correctness with Large Language Models (2608.15145)GPT-5.6 Soltechnique used95%
The Unwritten Benchmark: A New Challenge for Multimodal Machine Learning in Abstract Perceptual Reasoning (2608.14558)Gemini 3.7 Flashevaluates98%
The Unwritten Benchmark: A New Challenge for Multimodal Machine Learning in Abstract Perceptual Reasoning (2608.14558)GPT-5.6 Solevaluates98%
Large Language Models Show Metacognitive Sensitivity in Medical Reasoning (2608.14552)GPT-5.6 Solevaluates99%
MemoryLake on MemoryArena: A Matched Study of Agent Memory Backends (2608.13883)GPT-5.6 Soltechnique used95%
From Passive Delegates to Strategic Negotiators: Reinforcing Social Reasoning in Small Language Models with SocialRL (2608.13787)GPT-5.6 Solevaluates99%
Intern-S2-Mobius: Foundation Model with Decoupled Knowledge and Reasoning (2608.14290)Qwen 3.8 Maxtechnique used85%
Implementing Computational Law in Wolfram Language for the Governance of Artificial Intelligence (2608.13958)GPT-5.6 Solevaluates99%
Explanation Multiplicity: Circuit-Level Interpretability Evidence Does Not Survive Defensible Analytic Variation (2608.13754)GPT-5.6 Solevaluates98%
Wrong but Useful: Trajectory Value Beyond Answer Correctness in Multi-Agent Messages (2608.14375)Gemma 4evaluates95%
Wrong but Useful: Trajectory Value Beyond Answer Correctness in Multi-Agent Messages (2608.14375)GPT-5.6 Solevaluates98%
A Calibrated Test of Internal Action Maps: State Signals Without Global Affine Closure (2608.13626)Qwen 3.8 Maxevaluates98%
Depth-Aware Sensitivity Analysis of Mixture-of-Experts Models via Magnitude-Based Expert Masking (2608.13565)Qwen 3.8 Maxevaluates98%
OrchestraBench: Evaluating Multi-Agent Orchestration Failure Modes, Recovery, and Decomposition Quality (2608.05263)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates90%
NiyamAI - An Intent-Bound AI Agent with Cryptographically Verifiable Guardrails using Zero-Knowledge Proofs (2608.07167)GPT-5.6 Solevaluates99%
ε-MemEvo: Adaptive Cross-Task Memory Transfer for LLM Program Evolution (2608.12522)AlphaEvolvecites88%
ε-MemEvo: Adaptive Cross-Task Memory Transfer for LLM Program Evolution (2608.12522)GPT-5.6 Solevaluates98%
Don't Want Your LLM to Recommend Nuclear Strike? Try Asking It in Japanese (2608.12373)Gemini 3.7 Flashevaluates98%
Don't Want Your LLM to Recommend Nuclear Strike? Try Asking It in Japanese (2608.12373)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates98%
Privacy-Preserving RAG by Concealing Sensitive Information from External LLMs (2608.12675)Qwen 3.8 Maxevaluates98%
Rethinking Normalization Placement for LLMs: Post-Norm under Curriculum Depth Growing (2608.13156)Qwen 3.8 Maxevaluates98%
Explanatory Engagement Under Rare Anomalous Failure: Asymptotic Rarity in Model Behavior (or: The Asymptotic AI) (2608.13063)Mistral OCR 4evaluates99%
Explanatory Engagement Under Rare Anomalous Failure: Asymptotic Rarity in Model Behavior (or: The Asymptotic AI) (2608.13063)Qwen 3.8 Maxevaluates99%
Enhancing Virtual Agents through SLMs and Edge-Computing: An Exploratory Evaluation of Think and Memory Processes (2608.13420)Qwen 3.8 Maxevaluates98%
Reasoning Jury: Multi-Model Consensus for Evaluating Reasoning Traces (2608.12585)Gemini 3.7 Flashevaluates95%
Reasoning Jury: Multi-Model Consensus for Evaluating Reasoning Traces (2608.12585)claude-fable-5, claude-sonnet-5, claude-opus-5evaluates95%
Linkedclaude-fable-5, claude-sonnet-5, claude-opus-5GPT-5.6
agentic AIlarge language modelspublished Aug 24, 2026 · arXiv:2608.20844

TRACE: Agentic Catalog Enrichment with Multi-source Evidence Grounding

LLM agentscriterion revisionpublished Aug 24, 2026 · arXiv:2608.20729

Calibrating Criterion Revision in LLM Agents: Failure Modes and a Trace-Anchored Protocol

LinkedQwen 3.8 Max
instrumental convergenceagentic AIpublished Aug 24, 2026 · arXiv:2608.20940

The Logic of Machine Self-Preservation

agent skillscontinuous evaluationpublished Aug 24, 2026 · arXiv:2608.20614

Evaluating Skills, Not Just Agents: Agentic Continuous Evaluation of Skills

LLM agentsruntime interventionpublished Aug 24, 2026 · arXiv:2608.21027

Don't Solve, Just Compare: Tiny Advisors for Runtime Intervention in LLM Agents

software agentsagent memorypublished Aug 24, 2026 · arXiv:2608.20664

DreamBench-SWE: A Multi-Session Memory-Hygiene Benchmark for Software Agents

mobile agentsVLM-as-judgepublished Aug 24, 2026 · arXiv:2608.20797

Automated Trajectory Evaluation for Mobile Agents via Step-Level Consequence Reasoning and Aggregation

LinkedQwen 3.8 Max
multi-agent systemsdocument authoringpublished Aug 24, 2026 · arXiv:2608.20786

Structure for Reading, Prose for Writing: Asymmetric Structural Conditioning in Multi-Agent Document Authoring

agentic searchconformal predictionpublished Aug 24, 2026 · arXiv:2608.20771

CAS: Conformalized Agentic Search via Adaptive Retrieval and Policy Weighting

large language model agentsintegrated photonicspublished Aug 24, 2026 · arXiv:2608.20688

VortexChat: An agentic framework for autonomous multi-objective integrated photonic design

multi-agent LLMsconformal calibrationpublished Aug 24, 2026 · arXiv:2608.20564

Consilience: Conformally Calibrated Communication Control for Hidden-Profile Multi-Agent Reasoning

agentic LLMsModel Context Protocolpublished Aug 24, 2026 · arXiv:2608.20397

Nexus: Depth-Adaptive KV-Cache Splicing and Retrieval-Decoupled Tool Routing for Agentic LLMs on Unified Memory

LinkedQwen 3.8 Max
AI coding agentssoftware development lifecyclepublished Aug 24, 2026 · arXiv:2608.20341

SDAD: Spec-Driven Agentic Development for the AI-Native SDLC

LLM agentshierarchical memorypublished Aug 24, 2026 · arXiv:2608.20631

Weighted Memory Tree: Remembering What Matters for Long-Horizon LLM Agents

LinkedGemma 4Qwen 3.8 Max
federated learninglarge language modelspublished Aug 24, 2026 · arXiv:2608.20518

FL-MAESTRO: Multi-Agent LLM Orchestration for Resource-Constrained Federated Learning

terminal agentsAI agentspublished Aug 24, 2026 · arXiv:2608.20485

Terminal Agents: A Survey of AI Agents in Command-Line Environments

agentic memoryretrieval failurepublished Aug 24, 2026 · arXiv:2608.20400

When Retrieval Fails Before It Begins: Structurally Indirect Prerequisite Eviction as a Retention Failure in Agentic Memory

large language modelshallucinationpublished Aug 21, 2026 · arXiv:2608.19206

Hallucination as a Feature, not a Defect: Evaluating a multi-agent architecture to transform speculative language-model outputs into testable scientific hypotheses

conversational data collectionstated-preference surveyspublished Aug 21, 2026 · arXiv:2608.20320

An Agentic Approach for Active Data Collection, Travel Behavior Modeling, and Weather-Sensitive Demand Prediction

mid-trainingagentic tool usepublished Aug 21, 2026 · arXiv:2608.20314

MidTool: Mid-training Data Synthesis for Agentic Tool Use

LinkedQwen 3.8 Max
recursive self-improvementLLM agentspublished Aug 21, 2026 · arXiv:2608.20318

AI4AI-Bench: Benchmarking LLM Agents in Algorithmic Design for Recursive Self-Improvement

LLM agentsskill inductionpublished Aug 21, 2026 · arXiv:2608.20274

Break It Down, Pass It On: Cross-Task Skill Transfer in LLM Agents

LLM agentsskill selectionpublished Aug 21, 2026 · arXiv:2608.19993

Optimal Skill Selection for LLM Agents with Provable Bicriteria Guarantees

LLM agentsfinancial compliancepublished Aug 21, 2026 · arXiv:2608.19974

ReguSim: Evaluating LLM Agent Rule Grounding in Financial Compliance

LinkedGemini 3.7 FlashDeepSeek-V4-Flash-0731
agentic AIneuroimagingpublished Aug 21, 2026 · arXiv:2608.19902

Bringing analytic rigor to agentic AI for science: The Brain Researcher platform for neuroimaging data analysis

LLM agentsreinforcement learningpublished Aug 21, 2026 · arXiv:2608.19880

EnvHarness: Awakening Static Worlds for Agent Learning

LLM agentsworkflow verificationpublished Aug 21, 2026 · arXiv:2608.19861

PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agents

LinkedGemini 3.7 Flashclaude-fable-5, claude-sonnet-5, claude-opus-5GPT-5.6 Luna
agentic reinforcement learningpolicy optimizationpublished Aug 21, 2026 · arXiv:2608.19842

SAPO: Single-Rollout Autoregressive Policy Optimization for Agentic Reinforcement Learning

LinkedQwen 3.8 Max
multi-agent systemslong-term memorypublished Aug 21, 2026 · arXiv:2608.19701

Beyond Memory Majority: Latent-Source Reasoning for Multi-Agent Memory Arbitration

agent memorystate trackingpublished Aug 21, 2026 · arXiv:2608.19652

Can Agent Memory Systems Track Evolving State?

LinkedQwen 3.8 MaxDeepSeek-V4-Flash-0731
interactive AI agentsactive inferencepublished Aug 21, 2026 · arXiv:2608.19202

Active Inference as Context Acquisition for AI Agents

AI agentsscientific data servicespublished Aug 21, 2026 · arXiv:2608.19625

Scientific Data Skills: Enabling Agent-Ready Scientific Data Services at Scale

AI agentsscientific research automationpublished Aug 21, 2026 · arXiv:2608.19511

Trust via Auditable Records for Communities of AI Scientist Agents

AI agentstool-use safetypublished Aug 21, 2026 · arXiv:2608.19303

Outcome Monitors: Recovery Affordances for Silent Tool Failures

multi-agent systemslarge language modelspublished Aug 20, 2026 · arXiv:2608.18103

DeepTCM1.0: A Multi-Expert AI Agent for Deciphering Mechanisms of Chinese Herbal Formulae Based on General Large Language Models

LinkedDeepSeek V4-Flash
latent multi-agent communicationcovert coordinationpublished Aug 20, 2026 · arXiv:2608.19161

Beyond the Transcript: Detecting Covert Co ordination in Latent Multi-Agent Communication

LinkedQwen 3.8 Max
medical question answeringmulti-agent systemspublished Aug 20, 2026 · arXiv:2608.19029

Adaptive Memory and Reflection Multi-Agent System for Medical Question Answering

agentic AILLM-as-a-judgepublished Aug 20, 2026 · arXiv:2608.19002

A Theory of Post-hoc Debate Judgement

multi-turn reinforcement learningagentic workflowspublished Aug 20, 2026 · arXiv:2608.18682

RTPO: Reverse-Turn Policy Optimization for Stabilizing Agentic RL Training

generative AIlarge language modelspublished Aug 20, 2026 · arXiv:2608.18677

Sanyu Studio: A Multi-Agent System for Art-Historical Narrative Construction

terminal agentsagentic task synthesispublished Aug 20, 2026 · arXiv:2608.18580

FACET: Preserving Source Intent and Executable State in Terminal Task Synthesis

AI agentsmulti-agent systemspublished Aug 20, 2026 · arXiv:2608.18543

Bridging Search and CRM: Productionizing AI Product Research Agents for Customer Re-Engagement

AI code agentssoftware engineering agentspublished Aug 20, 2026 · arXiv:2608.18389

A Jagged Frontier: Evaluating Robustness of Code Agents to Semantics-Preserving Transformations

LinkedQwen 3.8 Maxclaude-fable-5, claude-sonnet-5, claude-opus-5
looped language modelscompositional tool callingpublished Aug 20, 2026 · arXiv:2608.18171

Looped Language Models Improve Compositional Tool Calling

agentic codingmulti-agent systemspublished Aug 20, 2026 · arXiv:2608.18167

Adversarial Review: Structured Disagreement for Grounded Agentic Code Review

AI reasoning agentschain-of-thoughtpublished Aug 20, 2026 · arXiv:2608.18078

Position: Collusion Risks Among AI Reasoning Agents Justify Certification Requirements for Making Market Decisions

LinkedDeepSeek V4-Flash
multi-agent systemsscientific discoverypublished Aug 20, 2026 · arXiv:2608.19047

Eureka: Task-Conditioned Meta-Agent Orchestration for Scientific Discovery

agentic AIreinforcement learningpublished Aug 20, 2026 · arXiv:2608.18852

SkillGate: Training In-Policy Skill Selection in Long-Horizon Agents

last-mile deliverymulti-agent decision-makingpublished Aug 20, 2026 · arXiv:2608.18846

ORBITER: Conflict-Aware Decision-Making for Agentic Last-Mile Delivery

agent evaluationautomatic metricspublished Aug 20, 2026 · arXiv:2608.18744

Metrics That Write Themselves: Evolving an Evaluator from Its Own Blind Spots

multi-agent systemsconversational business intelligencepublished Aug 20, 2026 · arXiv:2608.18740

A Multi-Agent Platform for Automated Enterprise Analytics and Insight Generation

cyber threat intelligenceLLM agentspublished Aug 20, 2026 · arXiv:2608.18613

CTIFoundry: An Agent-Native Corpus Scaffold for Cyber Threat Intelligence

Linkedclaude-fable-5, claude-sonnet-5, claude-opus-5
long-horizon agent evaluationmulti-agent competitionpublished Aug 20, 2026 · arXiv:2608.18423

FM-Bench: A Benchmark for Long-Horizon Management with Competing Agents

Linkedclaude-fable-5, claude-sonnet-5, claude-opus-5
computer-use agentsGUI groundingpublished Aug 20, 2026 · arXiv:2608.18307

ComponentBench: Diagnosing Component-Level Failures in Computer-Use Agents

LinkedQwen 3.8 MaxGemini 3.7 FlashGPT-5.6 Luna
agentic AIartificial intelligencepublished Aug 20, 2026 · arXiv:2608.18110

Emergence of Agentic AI: A Review on Evolution, Background, Working Principles, Applications, Adoption Factors, and Future Research Directions

large language model agentsself-evolving agentspublished Aug 20, 2026 · arXiv:2608.18104

Self-Evolving Agents as Dynamic Graph Transformation: A Survey and New Perspective

banking agentsadaptive fraudpublished Aug 20, 2026 · arXiv:2608.18136

FraudBench: Stress-Testing Policy-Grounded Banking Agents Against Adaptive Fraud

AI agentsinvestment managementpublished Aug 20, 2026 · arXiv:2608.18099

FinSkillBench: Evaluating AI Agents and Domain Skills for Investment Management

multi-agent systemslarge language modelspublished Aug 20, 2026 · arXiv:2608.18092

Position: Multi-Agent Systems Should Prioritize Concurrency Control

AI agentsbehavioral testingpublished Aug 20, 2026 · arXiv:2608.18081

Position: Behavioral Systems Require Behavioral Tests

AI agentsknowledge-work agentspublished Aug 19, 2026 · arXiv:2608.18050

StagedWorkspace: A Versioned Workspace for Knowledge-Work Agents

LinkedGPT-5.6 LunaGemini 3.7 Flash
AI agentslarge language modelspublished Aug 19, 2026 · arXiv:2608.17800

StartupBench: Benchmarking General-Purpose Agents on Market-Validated End-to-End Workflows

LLM agent memorypersistent memorypublished Aug 19, 2026 · arXiv:2608.17756

D²ACCI: A Dual-Loop Diagnostic Protocol for Evidence-Preserving Agent Memory

long-horizon agentsagent oversightpublished Aug 19, 2026 · arXiv:2608.17718

Beyond Suspicious Steps: Ontological Trust in Long-Horizon Agents

self-evolving agentsfinancial agentspublished Aug 19, 2026 · arXiv:2608.17684

Auditing Self-Evolution in Financial Agents: Capability Gains, Security Drift, and Execution-Interface Mismatch

LinkedQwen 3.8 Max
LLM agentsgroup polarizationpublished Aug 19, 2026 · arXiv:2608.17665

GraphWake: Group Polarization via Memory-Mediated Polarization Cascade in LLM-Agent Communities

AI agent safetyagent skillspublished Aug 19, 2026 · arXiv:2608.17588

TRUSS: Towards Task-Reliable and User-Safe Automated Agent Skill Generation

LinkedGPT-5.6 Luna
agentic reinforcement learningagent harnessespublished Aug 19, 2026 · arXiv:2608.17528

Agent Lightning v1.0: Towards Harnessed Agentic RL

LinkedQwen 3.8 Max
LLM agentsalgorithmic designpublished Aug 19, 2026 · arXiv:2608.17471

When AI Designs AI: Innovation or Imitation?

LLM agentsagent harnessespublished Aug 19, 2026 · arXiv:2608.17433

Task-Aware Harness Provisioning for LLM Agents in Mission-Critical Infrastructure Operations

browser agentsweb automationpublished Aug 19, 2026 · arXiv:2608.17319

Wuying-Browser-Agent: Real-World Centric Fundamental Long-Horizon Browser Agents

interactive reinforcement learningmulti-turn dialoguepublished Aug 19, 2026 · arXiv:2608.17499

Towards Better Agents for Multi-Turn User Interaction: The Next User Turn Is More Than Context

reinforcement learninggroup-relative policy optimizationpublished Aug 19, 2026 · arXiv:2608.17289

PlanPO: Group Planning-Aware Policy Optimization for Multi-Turn Agentic LLMs

Show more