Persistent decision history plus role-separated redacted evaluation lets agents continually improve skills across sessions and outperform commercial deep-research and prior skill-evolution baselines.
hub Canonical reference
arXiv preprint arXiv:2404.14387 , year=
Canonical reference. 100% of citing Pith papers cite this work as background.
20
Pith papers citing it
Background
100% of classified citations
hub tools
citation-role summary
background 6
citation-polarity summary
roles
background 6polarities
background 6representative citing papers
EvoTrainer co-evolves LLM policies and training harnesses via empirical feedback to match or exceed human-engineered RL on math reasoning, code generation, and long-horizon software engineering.
PACE coordinates low-risk prompt evolution with validated higher-risk control-logic updates to improve frozen SLM agents on benchmarks without model retraining.
EvoVid proposes a temporal-centric self-evolution framework for Video-LLMs that uses temporal-aware Questioner and temporal-grounded Solver rewards to improve performance directly from unannotated videos.
Mem-π is a framework using a dedicated model and decision-content decoupled RL to generate context-specific guidance on demand for LLM agents, outperforming retrieval baselines by over 30% on web navigation.
Solvita is an agentic evolution system using Planner, Solver, Oracle, and Hacker agents with trainable graph knowledge networks updated by reinforcement learning on pass/fail and vulnerability signals to achieve SOTA code generation performance.
Video-Zero is an annotation-free Questioner-Solver co-evolution framework that centers self-evolution on temporally localized evidence to improve video VLMs.
G-Zero uses the Hint-δ intrinsic reward to drive co-evolution between a Proposer and Generator via GRPO and DPO, providing a theoretical suboptimality guarantee for self-improvement from internal dynamics alone.
DarwinTOD proposes a dual-loop LLM-driven framework with an Evolvable Strategy Bank that enables lifelong autonomous improvement in task-oriented dialog systems through online multi-agent critique and offline evolutionary refinement.
Survey that defines agentic RL for LLMs via POMDPs, introduces a taxonomy of planning/tool-use/memory/reasoning capabilities and domains, and compiles open environments from over 500 papers.
Self-evolving LLM agents introduce persistent, amplifying security threats that static defenses cannot address, as shown by analysis of 25 attack surface cells and case studies.
Trace2Policy's EISR iteratively refines expert-derived rules into compiled Python code reaching 79.6% accuracy on skewed compliance tasks, outperforming one-shot LLM distillation and a deployed LLM baseline.
Existing methods for turning LLM interaction experience into parametric skills collapse over multiple iterations; principle-level experience, step-wise injection, and off-policy teacher distillation yield more stable continual learning.
D²Evo mines medium-difficulty anchors from the current model, trains a Questioner to generate matching questions, and jointly optimizes Solver and Questioner for progressive gains, outperforming baselines on math reasoning with under 2K real samples.
Bian Que is an agentic framework using a unified operational paradigm, flexible Skill Arrangement, and self-evolving mechanism to automate O&M tasks, achieving 75% alert reduction and over 50% MTTR cut in production deployment.
ACE introduces a solver-adversary loop where an LLM generates both candidate programs and adversarial tests, using execution outcomes for preference optimization to achieve 3-7% pass@1 gains on code benchmarks without ground-truth code.
SOLAR introduces a self-optimizing agent using meta-learning on model weights and RL-driven strategy discovery for lifelong adaptation in LLMs, claiming superior performance on reasoning tasks across domains.
Question generation produces a hidden-state signal that predicts final correctness before the answer is produced, yet gating interventions based on that signal do not reliably improve trajectories.
A survey consolidating frameworks, data practices, large action models, benchmarks, applications, and research gaps in LLM-brained GUI agents.
The paper reviews conceptual foundations, methodological innovations, effective designs, critical challenges, and future directions for LLM-based Agentic Reinforcement Learning.
citing papers explorer
-
SkillHone: A Harness for Continual Agent Skill Evolution Through Persistent Decision History
Persistent decision history plus role-separated redacted evaluation lets agents continually improve skills across sessions and outperform commercial deep-research and prior skill-evolution baselines.
-
EvoTrainer: Co-Evolving LLM Policies and Training Harnesses for Autonomous Agentic Reinforcement Learning
EvoTrainer co-evolves LLM policies and training harnesses via empirical feedback to match or exceed human-engineered RL on math reasoning, code generation, and long-horizon software engineering.
-
PACE: Two-Timescale Self-Evolution for Small Language Model Agents
PACE coordinates low-risk prompt evolution with validated higher-risk control-logic updates to improve frozen SLM agents on benchmarks without model retraining.
-
EvoVid: Temporal-Centric Self-Evolution for Video Large Language Models
EvoVid proposes a temporal-centric self-evolution framework for Video-LLMs that uses temporal-aware Questioner and temporal-grounded Solver rewards to improve performance directly from unannotated videos.
-
Mem-$\pi$: Adaptive Memory through Learning When and What to Generate
Mem-π is a framework using a dedicated model and decision-content decoupled RL to generate context-specific guidance on demand for LLM agents, outperforming retrieval baselines by over 30% on web navigation.
-
Solvita: Enhancing Large Language Models for Competitive Programming via Agentic Evolution
Solvita is an agentic evolution system using Planner, Solver, Oracle, and Hacker agents with trainable graph knowledge networks updated by reinforcement learning on pass/fail and vulnerability signals to achieve SOTA code generation performance.
-
Video-Zero: Self-Evolution Video Understanding
Video-Zero is an annotation-free Questioner-Solver co-evolution framework that centers self-evolution on temporally localized evidence to improve video VLMs.
-
G-Zero: Self-Play for Open-Ended Generation from Zero Data
G-Zero uses the Hint-δ intrinsic reward to drive co-evolution between a Proposer and Generator via GRPO and DPO, providing a theoretical suboptimality guarantee for self-improvement from internal dynamics alone.
-
DarwinTOD: LLM-driven Lifelong Self-evolution for Task-oriented Dialog Systems
DarwinTOD proposes a dual-loop LLM-driven framework with an Evolvable Strategy Bank that enables lifelong autonomous improvement in task-oriented dialog systems through online multi-agent critique and offline evolutionary refinement.
-
The Landscape of Agentic Reinforcement Learning for LLMs: A Survey
Survey that defines agentic RL for LLMs via POMDPs, introduces a taxonomy of planning/tool-use/memory/reasoning capabilities and domains, and compiles open environments from over 500 papers.
-
Safety in Self-Evolving LLM Agent Systems: Threats, Amplification, and Case Studies
Self-evolving LLM agents introduce persistent, amplifying security threats that static defenses cannot address, as shown by analysis of 25 attack surface cells and case studies.
-
Trace2Policy: From Expert Behavior Traces to Self-Evolving Decision Agents
Trace2Policy's EISR iteratively refines expert-derived rules into compiled Python code reaching 79.6% accuracy on skewed compliance tasks, outperforming one-shot LLM distillation and a deployed LLM baseline.
-
Rethinking Continual Experience Internalization for Self-Evolving LLM Agents
Existing methods for turning LLM interaction experience into parametric skills collapse over multiple iterations; principle-level experience, step-wise injection, and off-policy teacher distillation yield more stable continual learning.
-
D$^2$Evo: Dual Difficulty-Aware Self-Evolution for Data-Efficient Reinforcement Learning
D²Evo mines medium-difficulty anchors from the current model, trains a Questioner to generate matching questions, and jointly optimizes Solver and Questioner for progressive gains, outperforming baselines on math reasoning with under 2K real samples.
-
Bian Que: An Agentic Framework with Flexible Skill Arrangement for Online System Operations
Bian Que is an agentic framework using a unified operational paradigm, flexible Skill Arrangement, and self-evolving mechanism to automate O&M tasks, achieving 75% alert reduction and over 50% MTTR cut in production deployment.
-
ACE: Self-Evolving LLM Coding Framework via Adversarial Unit Test Generation and Preference Optimization
ACE introduces a solver-adversary loop where an LLM generates both candidate programs and adversarial tests, using execution outcomes for preference optimization to achieve 3-7% pass@1 gains on code benchmarks without ground-truth code.
-
SOLAR: A Self-Optimizing Open-Ended Autonomous Agent for Lifelong Learning and Continual Adaptation
SOLAR introduces a self-optimizing agent using meta-learning on model weights and RL-driven strategy discovery for lifelong adaptation in LLMs, claiming superior performance on reasoning tasks across domains.
-
What Am I Missing? Question-Answering as Hidden State Probing
Question generation produces a hidden-state signal that predicts final correctness before the answer is produced, yet gating interventions based on that signal do not reliably improve trajectories.
-
Large Language Model-Brained GUI Agents: A Survey
A survey consolidating frameworks, data practices, large action models, benchmarks, applications, and research gaps in LLM-brained GUI agents.
-
Rethinking Agentic Reinforcement Learning In Large Language Models
The paper reviews conceptual foundations, methodological innovations, effective designs, critical challenges, and future directions for LLM-based Agentic Reinforcement Learning.