Jianbo Wu

Di Zhang, Junxian Li, Jingdi Lei et al.

Vision-language models (VLMs) have shown remarkable advancements in multimodal reasoning tasks. However, they still often generate inaccurate or irrelevant responses due to issues like hallucinated image understandings or unrefined reasoning paths. To address these challenges, we introduce Critic-V, a novel framework inspired by the Actor-Critic paradigm to boost the reasoning capability of VLMs. This framework decouples the reasoning process and critic process by integrating two independent components: the Reasoner, which generates reasoning paths based on visual and textual inputs, and the Critic, which provides constructive critique to refine these paths. In this approach, the Reasoner generates reasoning responses according to text prompts, which can evolve iteratively as a policy based on feedback from the Critic. This interaction process was theoretically driven by a reinforcement learning framework where the Critic offers natural language critiques instead of scalar rewards, enabling more nuanced feedback to boost the Reasoner's capability on complex reasoning tasks. The Critic model is trained using Direct Preference Optimization (DPO), leveraging a preference dataset of critiques ranked by Rule-based Reward~(RBR) to enhance its critic capabilities. Evaluation results show that the Critic-V framework significantly outperforms existing methods, including GPT-4V, on 5 out of 8 benchmarks, especially regarding reasoning accuracy and efficiency. Combining a dynamic text-based policy for the Reasoner and constructive feedback from the preference-optimized Critic enables a more reliable and context-sensitive multimodal reasoning process. Our approach provides a promising solution to enhance the reliability of VLMs, improving their performance in real-world reasoning-heavy multimodal applications such as autonomous driving and embodied intelligence.

Ethan Smith, Rami Seid, Alberto Hojel et al.

Low-Rank Adaptation (LoRA) and other parameter-efficient fine-tuning (PEFT) methods provide low-memory, storage-efficient solutions for personalizing text-to-image models. However, these methods offer little to no improvement in wall-clock training time or the number of steps needed for convergence compared to full model fine-tuning. While PEFT methods assume that shifts in generated distributions (from base to fine-tuned models) can be effectively modeled through weight changes in a low-rank subspace, they fail to leverage knowledge of common use cases, which typically focus on capturing specific styles or identities. Observing that desired outputs often comprise only a small subset of the possible domain covered by LoRA training, we propose reducing the search space by incorporating a prior over regions of interest. We demonstrate that training a hypernetwork model to generate LoRA weights can achieve competitive quality for specific domains while enabling near-instantaneous conditioning on user input, in contrast to traditional training methods that require thousands of steps.

Di Zhang, Weida Wang, Junxian Li et al.

This paper target in addressing the challenges of underthinking and overthinking in long chain-of-thought (CoT) reasoning for Large Reasoning Models (LRMs) by introducing Reasoning Control Fields (RCF)--a novel test-time approach that injects structured control signals to guide reasoning from a tree search perspective. RCF enables models to adjust reasoning effort according to given control conditions when solving complex tasks. Additionally, we present the Control-R-4K dataset, which consists of challenging problems annotated with detailed reasoning processes and corresponding control fields. To further enhance reasoning control, we propose a Conditional Distillation Finetuning (CDF) method, which trains model--particularly Control-R-32B--to effectively adjust reasoning effort during test time. Experimental results on benchmarks such as AIME2024 and MATH500 demonstrate that our approach achieves state-of-the-art performance at the 32B scale while enabling a controllable Long CoT reasoning process (L-CoT). Overall, this work introduces an effective paradigm for controllable test-time scaling reasoning.