RL post-training composes primitive rewrite skills into reusable macro and parallel contraction strategies that solve problems inaccessible to the base model under large sampling budgets.
Emergent hierarchical reasoning in llms through reinforcement learning
13 Pith papers cite this work. Polarity classification is still indexing.
13
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
roles
background 2polarities
background 2representative citing papers
CARL trains a critic for segment-level credit assignment from binary outcomes in LLM tool-use trajectories, yielding 6.7-9.7 point accuracy gains and 53% fewer calls on solvable questions across five benchmarks.
Small 7B reasoning models were fine-tuned on synthetic and curated QFT problems using RL and SFT, yielding performance gains, error analysis, and public release of data and traces.
SFT supplies entangled compositional traces of atomic skills and routing modules; RL identifies those modules and enables recombination on novel compositions outside the SFT support.
GPRL carries a k-dimensional skew-symmetric preference structure into policy updates with per-dimension advantages and a drift monitor, yielding 56.51% length-controlled win rate on AlpacaEval 2.0 from Llama-3-8B-Instruct while outperforming SimPO and SPPO on other benchmarks.
G²TR reduces visual tokens and prefill compute by 1.94x in separate-encoder UMMs via generation-guided importance from VAE latent consistency, balanced selection, and merging, while preserving reasoning accuracy and editing quality.
S-trace adds sparse eligibility traces to RLVR that mask low-entropy tokens, outperforming GRPO by 0.49-3.16% pass@16 on Qwen3 models while improving sample and token efficiency.
Survey mapping RL techniques onto LLM training and highlighting gaps in value-based, off-policy, and bootstrapping methods.
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
Derives a token-level entropy change approximation revealing four factors, identifies limitations in prior entropy interventions, and proposes STEER which adaptively reweights tokens to mitigate collapse and improve performance on math and coding benchmarks.
For exponential family data, optimal FC-DNN parameters equal the RG fixed points of the input characteristic parameters, making training equivalent to RG coarse-graining.
Proposes Modality-Aware Credit Assignment (MoCA) with blindfolded-reasoning proxy to reward perception fidelity separately from reasoning in VLMs.
citing papers explorer
-
RL Post-Training Builds Compositional Reasoning Strategies
RL post-training composes primitive rewrite skills into reusable macro and parallel contraction strategies that solve problems inaccessible to the base model under large sampling budgets.
-
Knowing When to Ask: Segment-Level Credit Assignment for LLM Tool Use
CARL trains a critic for segment-level credit assignment from binary outcomes in LLM tool-use trajectories, yielding 6.7-9.7 point accuracy gains and 53% fewer calls on solvable questions across five benchmarks.
-
Fine-Tuning Small Reasoning Models for Quantum Field Theory
Small 7B reasoning models were fine-tuned on synthetic and curated QFT problems using RL and SFT, yielding performance gains, error analysis, and public release of data and traces.
-
From Reasoning Traces to Reusable Modules: Understanding Compositional Generalization in Language Model Reasoning
SFT supplies entangled compositional traces of atomic skills and routing modules; RL identifies those modules and enables recombination on novel compositions outside the SFT support.
-
General Preference Reinforcement Learning
GPRL carries a k-dimensional skew-symmetric preference structure into policy updates with per-dimension advantages and a drift monitor, yielding 56.51% length-controlled win rate on AlpacaEval 2.0 from Llama-3-8B-Instruct while outperforming SimPO and SPPO on other benchmarks.
-
G$^2$TR: Generation-Guided Visual Token Reduction for Separate-Encoder Unified Multimodal Models
G²TR reduces visual tokens and prefill compute by 1.94x in separate-encoder UMMs via generation-guided importance from VAE latent consistency, balanced selection, and merging, while preserving reasoning accuracy and editing quality.
-
Beyond Uniform Credit Assignment: Selective Eligibility Traces for RLVR
S-trace adds sparse eligibility traces to RLVR that mask low-entropy tokens, outperforming GRPO by 0.49-3.16% pass@16 on Qwen3 models while improving sample and token efficiency.
-
Modularized Reinforcement Learning on LLMs: From MDP Creation to Exploration and Learning
Survey mapping RL techniques onto LLM training and highlighting gaps in value-based, off-policy, and bootstrapping methods.
-
Trust Region On-Policy Distillation
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
-
Rethinking Entropy Interventions in RLVR: An Entropy Change Perspective
Derives a token-level entropy change approximation revealing four factors, identifies limitations in prior entropy interventions, and proposes STEER which adaptively reweights tokens to mitigate collapse and improve performance on math and coding benchmarks.
-
Interpreting FCDNNs via RG on Exponential Family
For exponential family data, optimal FC-DNN parameters equal the RG fixed points of the input characteristic parameters, making training equivalent to RG coarse-graining.
-
Bad Seeing or Bad Thinking? Rewarding Perception for Multimodal Reasoning
Proposes Modality-Aware Credit Assignment (MoCA) with blindfolded-reasoning proxy to reward perception fidelity separately from reasoning in VLMs.
- From Reasoning to Agentic: Credit Assignment in Reinforcement Learning for Large Language Models