mlsanityData to 7 Oct 2026

LLM agents and tool use

16 directions · 15,835 papers · 211 this week

Busiest this week

Most new since Mon 5 Oct

Software engineering coding agents

2,041 papers32 this weekMedium confidence in this grouping: mostly one topic, with blurry edgesMedium confidence in this grouping: mostly one topic, with blurry edges · About

  1. Cross-Provider Review as a Runtime Contract for Coding Agents: A Controlled Pilot and Fault-Injection Study
  2. The Harness as the Only Mutable Surface: Compliance-Bounded Self-Evolution of LLM Agents in Credit Pipelines, with a Measured Admission Gate
  3. Agent4RE: A Self-Refining Multi-agent Framework for End-to-End Software Requirements Engineering and Benchmarking
  4. TestJack: Should you trust the results in coding benchmarks? Agentic Coding Benchmarks Auditing via Evaluator Evolution
  5. Code Understanding is a Bottleneck for Coding Agents
Open →

All directions

Largest first · week from Mon 5 Oct

  1. General LLM agent frameworks and benchmarksagents, agent, language, llmLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About2,91719Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  2. Agent threats, guardrails and governanceagents, security, attack, agentMedium confidence in this grouping: mostly one topic, with blurry edges · About2,09228Medium confidence in this grouping: mostly one topic, with blurry edges
  3. Software engineering coding agentscode, software, agents, agentMedium confidence in this grouping: mostly one topic, with blurry edges · About2,04132Medium confidence in this grouping: mostly one topic, with blurry edges
  4. Web, search, and research agentsagents, research, search, webLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About1,67114Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  5. Multimodal and embodied agentsvisual, reasoning, multimodal, frameworkLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About1,46919Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  6. Social simulation and multi-agent debateagents, social, multi-agent, llmLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About1,0669Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  7. Medical and scientific domain agentsclinical, medical, reasoning, frameworkLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About9029Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  8. Reinforcement learning for LLM agentsreinforcement, reinforcement learning, training, agentsMedium confidence in this grouping: mostly one topic, with blurry edges · About85016Medium confidence in this grouping: mostly one topic, with blurry edges
  9. Agent memory systemsmemory, agents, retrieval, contextMedium confidence in this grouping: mostly one topic, with blurry edges · About71819Medium confidence in this grouping: mostly one topic, with blurry edges
  10. GUI and mobile agentsgui, gui agents, agents, mobileHigh confidence in this grouping: the papers form one clear topic · About5042High confidence in this grouping: the papers form one clear topic
  11. Tool learning and function callingtool, tools, llms, languageMedium confidence in this grouping: mostly one topic, with blurry edges · About4567Medium confidence in this grouping: mostly one topic, with blurry edges
  12. Financial trading and forecasting agentsfinancial, agents, trading, marketLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About3993Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  13. Agent skill learning and evolutionskill, skills, agent, agentsMedium confidence in this grouping: mostly one topic, with blurry edges · About25910Medium confidence in this grouping: mostly one topic, with blurry edges
  14. Educational agents and math problem solvingeducational, student, proof, mathematicalLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About2332Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  15. Agentic recommender systemsrecommendation, recommender, user, recommender systemsHigh confidence in this grouping: the papers form one clear topic · About1401High confidence in this grouping: the papers form one clear topic
  16. Agentic AI for wireless networksnetwork, networks, agentic, wirelessMedium confidence in this grouping: mostly one topic, with blurry edges · About1183Medium confidence in this grouping: mostly one topic, with blurry edges
18 new this week aren’t in a direction yet
  1. On the Clock: Towards Punctual and Productive Time-Budgeted AI Agents
  2. SciExam for ENSO: Can AI Agents Build Climate Models?
  3. RunningTab: Direct Workspace Interaction with Environment-Side Tabs
  4. Correct Answers, Unsupported Findings: Evidence Binding in Forensic Reconstruction of LLM Agent Logs
  5. Which Buildings Are Artificial Intelligence-Ready? A Measurement-Based Assessment Framework for AI Question Answering and Actuation
  6. Recursive Game Creator: An Agentic Product-Level Experience-Oriented Game Harness
  7. SpeedrunBench: Challenging LLM Agents with Video Game Speedrunning
  8. SIGMA: Self-Improving Alignment Generalization from a Model Spec
  9. ShanLiangRen: A Nutrition Agent for Personalized Daily Meal Planning
  10. A self-learning scientific agent for X-ray diffraction
  11. Unanimously Wrong: Certified Abstention from How Medical LLM Consensus Forms
  12. Decoupling What from Where: How Should a Small GUI Grounding Model Receive the Action Type?
  13. Agentic Design Space Exploration for Joint Hardware Configuration Selection and Mapping of AI Inference Workloads on Heterogeneous Edge SoCs
  14. Valid Stopping in Adaptive Generator-Verifier Loops
  15. Agentic schema-guided extraction of materials process knowledge from scientific literature
  16. EORestore-Agent: Fidelity-Guided Agentic Restoration of Remote Sensing Images with Composite Degradations
  17. Parallelism or Concession? Concurrency-Aware Procurement Negotiation for Agentic Commerce
  18. Process Constitutions and Process Stewards: Towards the Next Generation of BPM for Agentic Organizations

Weekly activity

Papers per week, 22 Jun – 5 Oct (last week partial)

  1. DirectionConfidencepapers16 wks
  2. Software engineering coding agentsMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 34; 29 Jun: 57; 6 Jul: 49; 13 Jul: 47; 20 Jul: 30; 27 Jul: 59; 3 Aug: 60; 10 Aug: 43; 17 Aug: 44; 24 Aug: 66; 31 Aug: 42; 7 Sep: 35; 14 Sep: 41; 21 Sep: 66; 28 Sep: 82; 5 Oct: 322,041787
  3. Agent threats, guardrails and governanceMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 42; 29 Jun: 37; 6 Jul: 40; 13 Jul: 30; 20 Jul: 30; 27 Jul: 46; 3 Aug: 55; 10 Aug: 35; 17 Aug: 37; 24 Aug: 47; 31 Aug: 47; 7 Sep: 32; 14 Sep: 49; 21 Sep: 50; 28 Sep: 62; 5 Oct: 282,092667
  4. General LLM agent frameworks and benchmarksLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 48; 29 Jun: 47; 6 Jul: 31; 13 Jul: 34; 20 Jul: 22; 27 Jul: 46; 3 Aug: 36; 10 Aug: 39; 17 Aug: 34; 24 Aug: 36; 31 Aug: 56; 7 Sep: 26; 14 Sep: 38; 21 Sep: 37; 28 Sep: 60; 5 Oct: 192,917609
  5. Multimodal and embodied agentsLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 17; 29 Jun: 38; 6 Jul: 20; 13 Jul: 21; 20 Jul: 13; 27 Jul: 13; 3 Aug: 33; 10 Aug: 25; 17 Aug: 19; 24 Aug: 26; 31 Aug: 29; 7 Sep: 18; 14 Sep: 28; 21 Sep: 30; 28 Sep: 53; 5 Oct: 191,469402
  6. Reinforcement learning for LLM agentsMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 14; 29 Jun: 12; 6 Jul: 14; 13 Jul: 11; 20 Jul: 8; 27 Jul: 13; 3 Aug: 17; 10 Aug: 14; 17 Aug: 15; 24 Aug: 13; 31 Aug: 17; 7 Sep: 10; 14 Sep: 11; 21 Sep: 31; 28 Sep: 57; 5 Oct: 16850273
  7. Web, search, and research agentsLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 14; 29 Jun: 21; 6 Jul: 26; 13 Jul: 28; 20 Jul: 21; 27 Jul: 31; 3 Aug: 27; 10 Aug: 30; 17 Aug: 21; 24 Aug: 28; 31 Aug: 27; 7 Sep: 30; 14 Sep: 23; 21 Sep: 19; 28 Sep: 41; 5 Oct: 141,671401
  8. Agent memory systemsMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 19; 29 Jun: 18; 6 Jul: 9; 13 Jul: 15; 20 Jul: 10; 27 Jul: 22; 3 Aug: 29; 10 Aug: 15; 17 Aug: 17; 24 Aug: 18; 31 Aug: 18; 7 Sep: 16; 14 Sep: 11; 21 Sep: 24; 28 Sep: 40; 5 Oct: 19718300
  9. Social simulation and multi-agent debateLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 13; 29 Jun: 11; 6 Jul: 14; 13 Jul: 11; 20 Jul: 9; 27 Jul: 16; 3 Aug: 21; 10 Aug: 11; 17 Aug: 13; 24 Aug: 16; 31 Aug: 18; 7 Sep: 13; 14 Sep: 16; 21 Sep: 19; 28 Sep: 25; 5 Oct: 91,066235
  10. Agent skill learning and evolutionMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 5; 29 Jun: 9; 6 Jul: 7; 13 Jul: 6; 20 Jul: 3; 27 Jul: 10; 3 Aug: 21; 10 Aug: 10; 17 Aug: 10; 24 Aug: 8; 31 Aug: 15; 7 Sep: 8; 14 Sep: 6; 21 Sep: 16; 28 Sep: 31; 5 Oct: 10259175
  11. Medical and scientific domain agentsLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 10; 29 Jun: 12; 6 Jul: 12; 13 Jul: 10; 20 Jul: 8; 27 Jul: 18; 3 Aug: 15; 10 Aug: 14; 17 Aug: 20; 24 Aug: 14; 31 Aug: 16; 7 Sep: 10; 14 Sep: 11; 21 Sep: 15; 28 Sep: 20; 5 Oct: 9902214
  12. GUI and mobile agentsHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 12; 29 Jun: 7; 6 Jul: 1; 13 Jul: 9; 20 Jul: 6; 27 Jul: 10; 3 Aug: 8; 10 Aug: 6; 17 Aug: 4; 24 Aug: 10; 31 Aug: 9; 7 Sep: 6; 14 Sep: 7; 21 Sep: 9; 28 Sep: 21; 5 Oct: 2504127
  13. Financial trading and forecasting agentsLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 7; 29 Jun: 5; 6 Jul: 8; 13 Jul: 4; 20 Jul: 7; 27 Jul: 7; 3 Aug: 9; 10 Aug: 8; 17 Aug: 5; 24 Aug: 9; 31 Aug: 14; 7 Sep: 3; 14 Sep: 6; 21 Sep: 17; 28 Sep: 13; 5 Oct: 3399125
  14. Tool learning and function callingMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 4; 29 Jun: 9; 6 Jul: 3; 13 Jul: 8; 20 Jul: 3; 27 Jul: 10; 3 Aug: 8; 10 Aug: 5; 17 Aug: 8; 24 Aug: 11; 31 Aug: 10; 7 Sep: 7; 14 Sep: 7; 21 Sep: 6; 28 Sep: 17; 5 Oct: 7456123
  15. Agentic AI for wireless networksMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 1; 29 Jun: 3; 6 Jul: 4; 13 Jul: 2; 20 Jul: 1; 27 Jul: 0; 3 Aug: 1; 10 Aug: 0; 17 Aug: 2; 24 Aug: 4; 31 Aug: 1; 7 Sep: 1; 14 Sep: 6; 21 Sep: 2; 28 Sep: 3; 5 Oct: 311834
  16. Agentic recommender systemsHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 0; 29 Jun: 2; 6 Jul: 4; 13 Jul: 0; 20 Jul: 5; 27 Jul: 5; 3 Aug: 5; 10 Aug: 2; 17 Aug: 2; 24 Aug: 2; 31 Aug: 1; 7 Sep: 2; 14 Sep: 2; 21 Sep: 4; 28 Sep: 5; 5 Oct: 114042
  17. Educational agents and math problem solvingLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 2; 29 Jun: 3; 6 Jul: 8; 13 Jul: 7; 20 Jul: 1; 27 Jul: 2; 3 Aug: 3; 10 Aug: 2; 17 Aug: 1; 24 Aug: 0; 31 Aug: 2; 7 Sep: 2; 14 Sep: 3; 21 Sep: 1; 28 Sep: 4; 5 Oct: 223343