ActFocus resolves the action bottleneck in agentic RL by reweighting token gradients toward action tokens using observed reward variance and an energy-based uncertainty term, outperforming PPO and GRPO by up to 65 percentage points.
Lmrl gym: Benchmarks for multi-turn reinforcement learning with language models
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 2
dataset 1
citation-polarity summary
polarities
background 3representative citing papers
JigsawRL achieves up to 1.85x higher throughput in LLM RL pipelines via pipeline multiplexing, sub-stage graphs, and look-ahead scheduling compared to prior systems.
SpeciaRL applies a dynamic verifier-based reward in reinforcement learning to steer reasoning LMMs toward correct and specific predictions on fine-grained open-world image classification tasks.
Visual-RFT applies reinforcement learning with verifiable perception rewards to improve large vision-language models on fine-grained classification, few-shot detection, and grounding tasks.
This survey categorizes agentic environments for LLMs by eight attributes and domains, introduces symbolic and neural synthesis paradigms with evaluation, and outlines four agent evolution pathways plus three environment evolution paradigms.
Sequential SFT followed by RL, guided by the Plasticity-Ceiling Framework, achieves higher performance ceilings in LLM mathematical reasoning than synchronized methods by optimizing data scale and transition timing.
citing papers explorer
-
Resolving Action Bottleneck: Agentic Reinforcement Learning Informed by Token-Level Energy
ActFocus resolves the action bottleneck in agentic RL by reweighting token gradients toward action tokens using observed reward variance and an energy-based uncertainty term, outperforming PPO and GRPO by up to 65 percentage points.
-
JigsawRL: Assembling RL Pipelines for Efficient LLM Post-Training
JigsawRL achieves up to 1.85x higher throughput in LLM RL pipelines via pipeline multiplexing, sub-stage graphs, and look-ahead scheduling compared to prior systems.
-
Specificity-aware reinforcement learning for fine-grained open-world classification
SpeciaRL applies a dynamic verifier-based reward in reinforcement learning to steer reasoning LMMs toward correct and specific predictions on fine-grained open-world image classification tasks.
-
Visual-RFT: Visual Reinforcement Fine-Tuning
Visual-RFT applies reinforcement learning with verifiable perception rewards to improve large vision-language models on fine-grained classification, few-shot detection, and grounding tasks.
-
Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application
This survey categorizes agentic environments for LLMs by eight attributes and domains, introduces symbolic and neural synthesis paradigms with evaluation, and outlines four agent evolution pathways plus three environment evolution paradigms.
-
Rethinking Expert Trajectory Utilization in LLM Post-training for Mathematical Reasoning
Sequential SFT followed by RL, guided by the Plasticity-Ceiling Framework, achieves higher performance ceilings in LLM mathematical reasoning than synchronized methods by optimizing data scale and transition timing.
- Alignment has a Fantasia Problem