paper-with-me

Papers

Tiny Inference-Time Scaling with Latent Verifiers

2026-03-23 · Davide Bucciarelli, Evelyn Turri, Lorenzo Baraldi, Marcella Cornia, Lorenzo Baraldi, Rita Cucchiara arxiv

Inference-time scaling has emerged as an effective way to improve generative models at test time by using a verifier to score and select candidate outputs. A common choice is to employ Multimodal Large Language Models (MLLMs) as verifiers, which can improve performance but introduce substantial inference-time cost. Indeed, diffusion pipelines operate in an autoencoder latent space to reduce computation, yet MLLM verifiers still require decoding candidates to pixel space and re-encoding them into the visual embedding space, leading to redundant and costly operations. In this work, we propose Verifier on Hidden States (VHS), a verifier that operates directly on intermediate hidden representations of Diffusion Transformer (DiT) single-step generators. VHS analyzes generator features without decoding to pixel space, thereby reducing the per-candidate verification cost while improving or matching the performance of MLLM-based competitors. We show that, under tiny inference budgets with only a small number of candidates per prompt, VHS enables more efficient inference-time scaling reducing joint generation-and-verification time by 63.3%, compute FLOPs by 51% and VRAM usage by 14.5% with respect to a standard MLLM verifier, achieving a +2.7% improvement on GenEval at the same inference-time budget.

📄 PDF Abstract BibTeX arXiv:2603.22492

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Trust but Verify! A Survey on Verification Design for Test-time Scaling

2025-08-20 · V Venktesh, Mandeep Rathee, Avishek Anand arxiv

Test-time scaling (TTS) has emerged as a new frontier for scaling the performance of Large Language Models. In test-time scaling, by using more computational resources during inference, LLMs can improve their reasoning p…

Multi-Agent Verification: Scaling Test-Time Compute with Multiple Verifiers

2025-02-27 · Shalev Lifshitz, Sheila A. McIlraith, Yilun Du

By utilizing more computational resources at test-time, large language models (LLMs) can improve without additional training. One common strategy uses verifiers to evaluate candidate outputs. In this work, we propose a n…

Llasa: Scaling Train-Time and Inference-Time Compute for Llama-based Speech Synthesis

2025-02-06 · Zhen Ye, Xinfa Zhu, Chi-Min Chan, Xinsheng Wang 외

Recent advances in text-based large language models (LLMs), particularly in the GPT series and the o1 model, have demonstrated the effectiveness of scaling both training-time and inference-time compute. However, current …

Speech Synthesis

Inference-Time Scaling for Complex Tasks: Where We Stand and What Lies Ahead

2025-03-31 · Vidhisha Balachandran, Jingya Chen, Lingjiao Chen, Shivam Garg 외

Inference-time scaling can enhance the reasoning capabilities of large language models (LLMs) on complex problems that benefit from step-by-step problem solving. Although lengthening generated scratchpads has proven effe…

MathSpatial Reasoning

Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning

2026-05-20 · Benhao Huang, Zhengyang Geng, Zico Kolter arxiv

Scaling test-time compute by iteratively updating a latent state has emerged as a powerful paradigm for reasoning. Yet the internal mechanisms that enable these iterative models to generalize beyond memorized patterns re…