Solve-Detect-Verify: Inference-Time Scaling with Flexible Generative Verifier (2026.acl-long)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models (LLMs) have enhanced capabilities in complex reasoning through step-by-step trace generation. |
| Approach: | They propose a generative verifier that dynamically allocates compute between rapid fast thinking and deliberative slow thinking. |
| Outcome: | The proposed solution outperforms GenPRM-32B on ProcessBench while requiring 2.3x fewer TFLOPS and 15x less training data. |
Similar Papers
Dyve: Thinking Fast and Slow for Dynamic Process Verification (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing process verification methods struggle with reliably assessing incomplete reasoning traces and are limited by the cost of high-quality human annotations or the inherent noise in automatically generated labels. |
| Approach: | They propose a dynamic process verifier that integrates fast and slow thinking to enhance reasoning error detection in large language models. |
| Outcome: | The proposed system outperforms existing process-based verifiers and maintains computational efficiency while maintaining high performance. |
MARS: Unleashing the Power of Speculative Decoding via Margin-Aware Verification (2026.findings-acl)
Copied to clipboard
Jingwei Song, Xinyu Wang, Hanbin Wang, Xiaoxuan Lei, Tianyu Shi, Shixin Han, Eric Yang, Xiao-Wen Chang, Lynn Ai
| Challenge: | Autoregressive large language models suffer from high inference latency due to memorybandwidth constraints. |
| Approach: | They propose a method that decouples generation and verification by decoupling tokens and a lightweight draft model. |
| Outcome: | The proposed method delivers consistent and significant speedups over state-of-the-art baselines while preserving generation quality across diverse benchmarks. |
Speculative Verification: Exploiting Information Gain for Speculative Decoding (2026.findings-acl)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are used for many applications but their size and computational cost make inference serving a significant challenge. |
| Approach: | They propose an efficient augmentation to Speculative Decoding (SD) that predicts speculation accuracy and dynamically adapts the verification length to maximize throughput. |
| Outcome: | The proposed model reduces wasted verification on rejected tokens and improves decoding efficiency. |
Think Right, Not More: Test-Time Scaling for Numerical Claim Verification (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Fact-checking real-world claims requires multistep reasoning and numerical reasoning . large language models are unable to understand nuance of numerical aspects . |
| Approach: | They propose scaling test-time compute (TTS) for large language models to solve this problem . they train a verifier model to navigate the space of possible reasoning paths . |
| Outcome: | The proposed approach achieves 1.8x higher efficiency than standard TTS while delivering 18.8% performance improvement over single-shot verification methods. |
PRIME: A Process-Outcome Alignment Benchmark for Verifiable Reasoning in Mathematics and Engineering (2026.acl-long)
Copied to clipboard
Xiangfeng Wang, Hangyu Guo, Yanlin Lai, Mitt Huang, Liang Zhao, Chengyuan Yao, Yinmin Zhang, Qi Han, null Xiaoxiaoren, Chun Yuan, Tong Xu, Zheng Ge, Xiangyu Zhang, Daxin Jiang
| Challenge: | Current outcome-centric verification paradigms neglect potential errors in the derivation process. |
| Approach: | They propose a process-aware RLVR training paradigm utilizing verifiers selected via **PRIME**. |
| Outcome: | The proposed approach outperforms the baseline verification paradigm on AIME24, AIME25, and Beyond-AIME models. |
Hard2Verify: A Step-Level Verification Benchmark for Open-Ended Frontier Math (2026.acl-long)
Copied to clipboard
| Challenge: | Large language model (LLM)-based reasoning systems have recently achieved gold medal-level performance in the IMO 2025 competition . |
| Approach: | They propose a human-annotated step-level verification benchmark that measures step- level verifiers at the frontier. |
| Outcome: | The proposed benchmark outperforms closed-source models in step-level verification and the impact of scaling verifier compute. |
VerifiAgent: a Unified Verification Agent in Language Model Reasoning (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing verification methods are typically model-specific or domain-restricted . Existing methods require significant computational resources and lack scalability . |
| Approach: | a unified verification agent integrates two levels of verification: meta-verification and tool-based adaptive verification. |
| Outcome: | The proposed agent outperforms baseline verification methods among reasoning tasks. |
ConfSpec: Efficient Step-Level Speculative Reasoning via Confidence-Gated Verification (2026.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to chain-of-thought reasoning incur high inference latency due to long generation traces. |
| Approach: | They propose a confidence-gated cascaded verification framework that reduces the trade-off between generation and verification. |
| Outcome: | The proposed framework achieves 2.24 speedups while matching target-model accuracy. |
From Data-Centric to Sample-Centric: Enhancing LLM Reasoning via Progressive Optimization (2026.acl-long)
Copied to clipboard
| Challenge: | Reinforcement learning with verifiable rewards (RLVR) has recently advanced the reasoning capabilities of large language models (LLMs). |
| Approach: | They propose a method that incorporates partial solution prefixes from expert demonstrations to guide the policy. |
| Outcome: | The proposed methods outperform strong baselines, yielding faster convergence and a higher performance ceiling. |
Tool learning via Inference-time Scaling and Cycle Verifier (2025.findings-acl)
Copied to clipboard
| Challenge: | In inference-time scaling, Chain-of-Thought (CoT) data is scarce or even unavailable. |
| Approach: | They propose a method which establishes an inference cycle to synthesize user queries and CoT data. |
| Outcome: | The proposed method achieves a 75.4% pass rate and a 79.6% win rate using small models in StableToolBench. |