REVIEW 13 cited by
AlphaMath Almost Zero: Process Supervision without Process
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
AlphaMath Almost Zero: Process Supervision without Process
read the original abstract
Although recent advancements in large language models (LLMs) have significantly improved their performance on various tasks, they still face challenges with complex and symbolic multi-step reasoning, particularly in mathematical reasoning. To bolster the mathematical reasoning capabilities of LLMs, most existing efforts concentrate on seeking assistance from either domain experts or GPT-4 for high-quality process-supervised data, which is not only expensive but also labor-intensive. In our study, we propose an innovative framework, AlphaMath, that bypasses the need for process annotations (from humans or GPTs) by leveraging Monte Carlo Tree Search (MCTS). This framework focuses on unleashing the potential of a well-pretrained LLM to autonomously enhance its mathematical reasoning. Specifically, we integrate a value model with the LLM, automatically generating both process supervision and step-level evaluation signals in MCTS. Furthermore, we propose an efficient inference strategy, step-level beam search, where the value model is crafted to assist the policy model (i.e., LLM) in navigating more effective reasoning paths, rather than solely relying on prior probabilities. The experimental results on both in-domain and out-of-domain datasets demonstrate that even without GPT-4 or human-annotated process supervision, our AlphaMath framework achieves comparable or superior results to previous state-of-the-art methods.
Forward citations
Cited by 13 Pith papers
-
Training the Orchestrator: A Supervised Approach to End-to-End PDDL Planning with LLM Agents
HALO trains an orchestrator policy on verifier-approved refinement trajectories across 11 PDDL domains, matching GPT-5-mini success rates at roughly 45x lower orchestration cost and cutting LLM calls by 40-50%.
-
Step-by-Step Optimization-like Reasoning in LLMs over Expanding Search Spaces
Introduces OPT* tasks and two training regimes (solver-guided online policy optimization with rank-based reward shaping and search-based offline RL) plus a theoretical link between search success and information extra...
-
VeriSpace: Spatially Grounded Action Verification for Vision-Language-Action Models
VeriSpace is a 3D-aware action verifier that improves test-time action selection in VLA models by encoding scenes with visual and geometric information and reasoning over spatial relations and goal progress.
-
Efficient Process Reward Modeling via Contrastive Mutual Information
CPMI labels reasoning-step rewards by measuring how much each step boosts mutual information with the correct answer relative to hard negatives, cutting labeling cost by 84% and tokens by 98% while improving accuracy.
-
Output-Space Search: Targeting LLM Generations in a Frozen Encoder-Defined Output Space
Retrieval-grounded reinforcement learning makes an autoregressive LLM hit requested coordinates in a frozen encoder-defined PCA space, giving an outer loop a low-dimensional target to sweep or optimize.
-
Sticker-TTS: Learn to Utilize Historical Experience with a Sticker-driven Test-Time Scaling Framework
Sticker-TTS uses three collaborating models to distill and reuse compact summaries (stickers) of past reasoning attempts, improving math benchmark accuracy at a claimed equivalent inference cost.
-
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.
-
SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training
Reinforcement learning post-training enables generalization to unseen textual rule variants and visual changes in foundation models, while supervised fine-tuning primarily leads to memorization.
-
The Lessons of Developing Process Reward Models in Mathematical Reasoning
Monte Carlo data synthesis for PRMs underperforms LLM-judge and human methods, Best-of-N evaluations suffer from process-outcome misalignment and score inflation, and consensus filtering yields better PRMs with higher...
-
Step-DPO: Step-wise Preference Optimization for Long-chain Reasoning of LLMs
Step-DPO performs preference optimization on individual reasoning steps rather than complete answers, producing nearly 3% accuracy gains on MATH for 70B+ parameter models with 10K preference pairs.
-
A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence
The paper delivers the first systematic review of self-evolving agents, structured around what components evolve, when adaptation occurs, and how it is implemented.
-
From System 1 to System 2: A Survey of Reasoning Large Language Models
The survey organizes the shift of LLMs toward deliberate System 2 reasoning, covering model construction techniques, performance on math and coding benchmarks, and future research directions.
-
Towards Large Reasoning Models: A Survey of Reinforced Reasoning with Large Language Models
The paper surveys reinforced reasoning techniques for LLMs, covering automated data construction, learning-to-reason methods, and test-time scaling as steps toward Large Reasoning Models.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.