C3PO: Critical-Layer, Core-Expert, Collaborative Pathway Optimization for Test-Time Expert Re-Mixing

• URL: http://arxiv.org/abs/2504.07964v1
• Date: Thu, 10 Apr 2025 17:59:56 GMT
• Title: C3PO: Critical-Layer, Core-Expert, Collaborative Pathway Optimization for Test-Time Expert Re-Mixing
• Authors: Zhongyang Li, Ziyue Li, Tianyi Zhou,
• Abstract summary: Mixture-of-Experts (MoE) Large Language Models (LLMs) suffer from severely sub-optimal expert pathways.<n> naive expert selection learned from pretraining leaves a surprising 10-20% accuracy gap for improvement.<n>We develop a novel class of test-time optimization methods to re-weight or "re-mixing" the experts in different layers jointly for each test sample.
• Score: 21.119495676190127
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Mixture-of-Experts (MoE) Large Language Models (LLMs) suffer from severely sub-optimal expert pathways-our study reveals that naive expert selection learned from pretraining leaves a surprising 10-20% accuracy gap for improvement. Motivated by this observation, we develop a novel class of test-time optimization methods to re-weight or "re-mixing" the experts in different layers jointly for each test sample. Since the test sample's ground truth is unknown, we propose to optimize a surrogate objective defined by the sample's "successful neighbors" from a reference set of samples. We introduce three surrogates and algorithms based on mode-finding, kernel regression, and the average loss of similar reference samples/tasks. To reduce the cost of optimizing whole pathways, we apply our algorithms merely to the core experts' mixing weights in critical layers, which enjoy similar performance but save significant computation. This leads to "Critical-Layer, Core-Expert, Collaborative Pathway Optimization (C3PO)". We apply C3PO to two recent MoE LLMs and examine it on six widely-used benchmarks. It consistently improves the base model by 7-15% in accuracy and outperforms widely used test-time learning baselines, e.g., in-context learning and prompt/prefix tuning, by a large margin. Moreover, C3PO enables MoE LLMs with 1-3B active parameters to outperform LLMs of 7-9B parameters, hence improving MoE's advantages on efficiency. Our thorough ablation study further sheds novel insights on achieving test-time improvement on MoE.

Related papers

• Large Language Model Enhanced Particle Swarm Optimization for Hyperparameter Tuning for Deep Learning Models [2.3949320404005436]
  Particle Swarm Optimization and Large Language Models (LLMs) have been individually applied in optimization and deep learning. Our work addresses this gap by integrating LLMs into PSO to reduce model evaluations and improve convergence. Our method speeds up search space exploration by substituting underperforming particle placements with best suggestions.
  arXiv  Detail & Related papers  (2025-04-19T00:54:59Z)
• Review, Refine, Repeat: Understanding Iterative Decoding of AI Agents with Dynamic Evaluation and Selection [71.92083784393418]
  Inference-time methods such as Best-of-N (BON) sampling offer a simple yet effective alternative to improve performance.<n>We propose Iterative Agent Decoding (IAD) which combines iterative refinement with dynamic candidate evaluation and selection guided by a verifier.
  arXiv  Detail & Related papers  (2025-04-02T17:40:47Z)
• S$^2$R: Teaching LLMs to Self-verify and Self-correct via Reinforcement Learning [51.84977135926156]
  We introduce S$2$R, an efficient framework that enhances LLM reasoning by teaching models to self-verify and self-correct during inference. Our results demonstrate that Qwen2.5-math-7B achieves an accuracy improvement from 51.0% to 81.6%, outperforming models trained on an equivalent amount of long-CoT distilled data.
  arXiv  Detail & Related papers  (2025-02-18T13:40:22Z)
• Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning [27.991291785091736]
  This work studies the problem of large language model (LLM) unlearning, aiming to remove unwanted data influences.<n>Despite the increasing demand for unlearning, a technically-grounded optimization framework is lacking.<n>We propose a simple yet effective unlearning optimization framework, called SimNPO, showing that simplicity' in removing the reliance on a reference model benefits unlearning.
  arXiv  Detail & Related papers  (2024-10-09T17:58:12Z)
• Self-supervised Preference Optimization: Enhance Your Language Model with Preference Degree Awareness [27.43137305486112]
  We propose a novel Self-supervised Preference Optimization (SPO) framework, which constructs a self-supervised preference degree loss combined with the alignment loss. The results demonstrate that SPO can be seamlessly integrated with existing preference optimization methods to achieve state-of-the-art performance.
  arXiv  Detail & Related papers  (2024-09-26T12:37:26Z)
• Efficient Continual Pre-training by Mitigating the Stability Gap [68.49269649759005]
  We study the behavior of Large Language Models (LLMs) during continual pre-training. We propose three effective strategies to enhance LLM performance within a fixed compute budget. Our strategies improve the average medical task performance of the OpenLlama-3B model from 36.2% to 40.7% with only 40% of the original training budget.
  arXiv  Detail & Related papers  (2024-06-21T02:28:37Z)
• 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)
• Take the Bull by the Horns: Hard Sample-Reweighted Continual Training Improves LLM Generalization [165.98557106089777]
  A key challenge is to enhance the capabilities of large language models (LLMs) amid a looming shortage of high-quality training data. Our study starts from an empirical strategy for the light continual training of LLMs using their original pre-training data sets. We then formalize this strategy into a principled framework of Instance-Reweighted Distributionally Robust Optimization.
  arXiv  Detail & Related papers  (2024-02-22T04:10:57Z)
• Optimizer Amalgamation [124.33523126363728]
  We are motivated to study a new problem named Amalgamation: how can we best combine a pool of "teacher" amalgamations into a single "student" that can have stronger problem-specific performance? First, we define three differentiable mechanisms to amalgamate a pool of analyticals by gradient descent. In order to reduce variance of the process, we also explore methods to stabilize the process by perturbing the target.
  arXiv  Detail & Related papers  (2022-03-12T16:07:57Z)
• Descending through a Crowded Valley - Benchmarking Deep Learning Optimizers [29.624308090226375]
  In this work, we aim to replace these anecdotes, if not with a conclusive ranking, then at least with evidence-backed anecdotes. To do so, we perform an extensive, standardized benchmark of fifteen particularly popular deep learnings. Our open-sourced results are available as challenging and well-tuned baselines for more meaningful evaluations of novel optimization methods.
  arXiv  Detail & Related papers  (2020-07-03T08:19:36Z)

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.