DeepConf: Enhancing LLM Reasoning Through Confidence-Based Inference Methods
By
che_shr_cat
Crackling crust, pillowy middle. The kind of bagel that earns a second cup of coffee.
Summary
DeepConf is a novel test-time inference method that enhances Large Language Models' reasoning capabilities by using internal log-probabilities to derive localized confidence scores. The method operates in two modes: offline filtering of completed reasoning traces with confidence-weighted majority voting, and online mode that dynamically adjusts reasoning depth based on confidence thresholds. This approach improves reasoning accuracy without requiring additional computational resources or model fine-tuning.
Key quotes
· 4 pulledDeepConf leverages the model's internal log-probabilities to derive localized confidence scores
Instead of treating all generated reasoning paths equally
Operates in two modes: an offline mode that filters completed reasoning traces and applies confidence-weighted majority
Enhances the reasoning capabilities of Large Language Models (LLMs)
You might also wanna read
RTP-LLM: Alibaba's High-Performance Inference Engine for Large Language Model Deployment
This paper presents RTP-LLM, a high-performance inference engine developed by Alibaba for industrial-scale deployment of Large Language Mode
arxiv.org·14d ago
DeepSeek-V3.1: Open-Source Language Model with Hybrid Inference for Advanced Reasoning and Coding
DeepSeek-V3.1 is an open-source large language model that introduces hybrid inference with both 'Think' and 'Non-Think' modes, optimized for
Product Hunt·9mo ago
RICP: A Teacher-Student Framework for Retrieved In-Context Principles from Mistakes in LLMs
This paper introduces Retrieved In-Context Principles (RICP), a novel teacher-student framework for improving Large Language Models (LLMs) t
arxiv.org·16d ago
LK Losses: A New Training Objective to Optimize Acceptance Rate in Speculative Decoding for LLMs
This paper introduces LK losses, a novel training objective for speculative decoding in large language models (LLMs). Speculative decoding a
arxiv.org·10d ago
CoT-PoT Ensembling: Efficient LLM Reasoning with Self-Consistency from Just Two Samples
This paper introduces a hybrid ensembling approach called CoT-PoT that combines Chain-of-Thought (CoT) and Program-of-Thought (PoT) reasonin
arxiv.org·4d ago
Researchers Develop Method to Predict Real-Time Progress in Reasoning Language Models
This research paper investigates whether real-time progress prediction is feasible for reasoning language models that use long latent chains
arxiv.org·16d ago