Streaming Looking Ahead with Token-level Self-reward

• URL: http://arxiv.org/abs/2503.00029v1
• Date: Mon, 24 Feb 2025 22:35:53 GMT
• Title: Streaming Looking Ahead with Token-level Self-reward
• Authors: Hongming Zhang, Ruixin Hong, Dong Yu,
• Abstract summary: We propose a policy model with token-level self-reward modeling (TRM) capability to eliminate the need for external models and extra communication.<n>In addition, we propose a streaming-looking-ahead (SLA) algorithm to further boost search efficiency with better parallelization.<n>If we combine SLA with reinforcement fine-tuning techniques such as DPO, SLA achieves an overall win rate of 89.4%.
• Score: 50.699168440048716
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Autoregressive decoding algorithms that use only past information often cannot guarantee the best performance. Recently, people discovered that looking-ahead algorithms such as Monte Carlo Tree Search (MCTS) with external reward models (RMs) can significantly improve models' output by allowing them to think ahead and leverage future outputs and associated rewards to guide the current generation. Such techniques can help the reinforcement fine-tuning phase by sampling better trajectories and the inference phase by selecting the better output. However, their high computational cost limits their applications, especially in streaming scenarios. To address this issue, we propose equipping the policy model with token-level self-reward modeling (TRM) capability to eliminate the need for external models and extra communication. We name the new architecture as Reward Transformer. In addition, we propose a streaming-looking-ahead (SLA) algorithm to further boost search efficiency with better parallelization. Experiments show that SLA achieves an overall win rate of 79.7\% against the baseline greedy decoding algorithm on three general-domain datasets with a frozen policy model while maintaining streaming efficiency. If we combine SLA with reinforcement fine-tuning techniques such as DPO, SLA achieves an overall win rate of 89.4\%.

Related papers

• DARS: Dynamic Action Re-Sampling to Enhance Coding Agent Performance by Adaptive Tree Traversal [55.13854171147104]
  Large Language Models (LLMs) have revolutionized various domains, including natural language processing, data analysis, and software development. We present Dynamic Action Re-Sampling (DARS), a novel inference time compute scaling approach for coding agents. We evaluate our approach on SWE-Bench Lite benchmark, demonstrating that this scaling strategy achieves a pass@k score of 55% with Claude 3.5 Sonnet V2.
  arXiv  Detail & Related papers  (2025-03-18T14:02:59Z)
• MM-RLHF: The Next Step Forward in Multimodal LLM Alignment [59.536850459059856]
  We introduce MM-RLHF, a dataset containing $mathbf120k$ fine-grained, human-annotated preference comparison pairs.<n>We propose several key innovations to improve the quality of reward models and the efficiency of alignment algorithms.<n>Our approach is rigorously evaluated across $mathbf10$ distinct dimensions and $mathbf27$ benchmarks.
  arXiv  Detail & Related papers  (2025-02-14T18:59:51Z)
• COrAL: Order-Agnostic Language Modeling for Efficient Iterative Refinement [80.18490952057125]
  Iterative refinement has emerged as an effective paradigm for enhancing the capabilities of large language models (LLMs) on complex tasks. We propose Context-Wise Order-Agnostic Language Modeling (COrAL) to overcome these challenges. Our approach models multiple token dependencies within manageable context windows, enabling the model to perform iterative refinement internally.
  arXiv  Detail & Related papers  (2024-10-12T23:56:19Z)
• A Distribution-Aware Flow-Matching for Generating Unstructured Data for Few-Shot Reinforcement Learning [1.0709300917082865]
  We introduce a distribution-aware flow matching approach to generate synthetic unstructured data for few-shot reinforcement learning.<n>Our approach addresses key challenges in traditional model-based RL, such as overfitting and data correlation.<n>Results demonstrate that our method achieves stable convergence in terms of maximum Q-value while enhancing frame rates by 30% in the initial timestamps.
  arXiv  Detail & Related papers  (2024-09-21T15:50:59Z)
• ITERTL: An Iterative Framework for Fine-tuning LLMs for RTL Code Generation [9.409062607311528]
  Large language models (LLMs) have demonstrated excellent performance, inspiring researchers to explore their use in automating register transfer level (RTL) code generation. Existing approaches to fine-tune LLMs for RTL generation typically are conducted on fixed datasets. We introduce an iterative training paradigm named ITERTL to mitigate these issues. Our model outperforms GPT4 and state-of-the-art (SOTA) open-source models, achieving remarkable 53.8% pass@1 rate on VerilogEval-human benchmark.
  arXiv  Detail & Related papers  (2024-06-28T01:44:57Z)
• Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms [50.808123629394245]
  Direct Alignment Algorithms (DDAs) like Direct Preference Optimization have emerged as alternatives to the classical RLHF pipeline. This work formulates and formalizes the reward over-optimization or hacking problem for DAAs and explores its consequences across objectives, training regimes, and model scales.
  arXiv  Detail & Related papers  (2024-06-05T03:41:37Z)
• Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learning [55.96599486604344]
  We introduce an approach aimed at enhancing the reasoning capabilities of Large Language Models (LLMs) through an iterative preference learning process. We use Monte Carlo Tree Search (MCTS) to iteratively collect preference data, utilizing its look-ahead ability to break down instance-level rewards into more granular step-level signals. The proposed algorithm employs Direct Preference Optimization (DPO) to update the LLM policy using this newly generated step-level preference data.
  arXiv  Detail & Related papers  (2024-05-01T11:10:24Z)
• LMUFormer: Low Complexity Yet Powerful Spiking Model With Legendre Memory Units [5.830814457423021]
  Transformer models have demonstrated high accuracy in numerous applications but have high complexity and lack sequential processing capability. We show how architectural modifications to a recurrent model can help push its performance toward Transformer models. We present a spiking version of this architecture, which introduces the benefit of states within the patch embedding and channel mixer modules.
  arXiv  Detail & Related papers  (2024-01-20T01:10:18Z)
• Highly Parallel Autoregressive Entity Linking with Discriminative Correction [51.947280241185]
  We propose a very efficient approach that parallelizes autoregressive linking across all potential mentions. Our model is >70 times faster and more accurate than the previous generative method.
  arXiv  Detail & Related papers  (2021-09-08T17:28:26Z)
• Hyperparameter optimization with REINFORCE and Transformers [2.1404235519012076]
  Reinforcement Learning has yielded promising results for Neural Architecture Search (NAS) We demonstrate how its performance can be improved by using a simplified Transformer block to model the policy network.
  arXiv  Detail & Related papers  (2020-06-01T13:35:48Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.