A Survey on LLM Inference-Time Self-Improvement

Xiangjue Dong, Maria Teleki, James Caverlee
12/18/2024
cs.CL

No official code found. Best implementation: dongxiangjue/Awesome-LLM-Self-Improvement (112 stars, MIT, last commit 2024-12-24).

Abstract

Techniques that enhance inference through increased computation at test-time have recently gained attention. In this survey, we investigate the current state of LLM Inference-Time Self-Improvement from three different perspectives: Independent Self-improvement, focusing on enhancements via decoding or sampling methods; Context-Aware Self-Improvement, leveraging additional context or datastore; and Model-Aided Self-Improvement, achieving improvement through model collaboration. We provide a comprehensive review of recent relevant studies, contribute an in-depth taxonomy, and discuss challenges and limitations, offering insights for future research.

View on arXivView PDF

Code Implementations(1)

dongxiangjue/Awesome-LLM-Self-ImprovementCommunity100%
MIT

A curated list of awesome LLM Inference-Time Self-Improvement (ITSI, pronounced "itsy") papers from our recent survey: A Survey on Large Language Model Inference-Time Self-Improvement.

1126Dec 15, 20241 years agoMIT

Similar Papers

Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters

85% match2024

Let's (not) just put things in Context: Test-Time Training for Long-Context LLMs

84% match2025

Theoretical Modeling of LLM Self-Improvement Training Dynamics Through Solver-Verifier Gap

84% match2025

Mind the Gap: Examining the Self-Improvement Capabilities of Large Language Models

83% match2024

Enhancing LLM Reasoning via Critique Models with Test-Time and Training-Time Supervision

83% match1 repo552024

Mitigating Tail Narrowing in LLM Self-Improvement via Socratic-Guided Sampling

82% match1 repo92024

Cite this paper

@article{dong2024survey,
  title  = {A Survey on LLM Inference-Time Self-Improvement},
  author = {Xiangjue Dong and Maria Teleki and James Caverlee},
  year   = {2024},
  eprint = {2412.14352},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2412.14352v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2412.14352.svg)](https://codeofpaper.com/paper/2412.14352)

Discussion