13 papers
Do Rollups Self-Organize? Multi-Agent Reinforcement Learning in the EIP-4844 Blob Fee Market
Alessio Brini
EIP-4844 gave Ethereum rollups a separate fee dimension for data, governed by a multiplicative update rule that targets three ``blobs'' per block. Whether this target is reached de…
Correcting WHERE, Preserving HOW: Compositional Generalization for Vision-Language-Action Models via Referential Guidance
Yanyan Zhang, Disheng Liu, Xinpeng Li +8
While Vision-Language-Action (VLA) models enable flexible action generation, their generalization across diverse environmental elements, including manipulated objects, destinations…
Mechanistic Diagnostics of Spatial Lexical Bias in Multimodal Large Language Model Spatial Reasoning
Chuang Ma, Qianying Liu, Tomoyuki Obuchi +6
Multimodal large language models (MLLMs) remain unreliable on spatial multiple-choice questions, and their failures are often attributed to poorly attended visual information. We i…
AgentCE-Bench: Agent Configurable Evaluation with Scalable Horizons and Controllable Difficulty under Lightweight Environments
Wang Yang, Chaoda Song, Xinpeng Li +7
Existing Agent benchmarks suffer from two critical limitations: high environment interaction overhead (up to 41\% of total evaluation time) and imbalanced task horizon and difficul…
Path-Lock Expert: Separating Reasoning Mode in Hybrid Thinking via Architecture-Level Separation
Shouren Wang, Wang Yang, Chuang Ma +7
Hybrid-thinking language models expose explicit /think and /no_think modes, but current designs do not separate them cleanly. Even in /no_think mode, models often emit long and sel…
MOSAIC: Multi-Objective Slice-Aware Iterative Curation for Alignment
Yipu Dou, Wang Yang
We study how to allocate a fixed supervised fine-tuning budget when three objectives must be balanced at once: multi-turn safety alignment, low over-refusal on benign boundary quer…
When Domains Interact: Asymmetric and Order-Sensitive Cross-Domain Effects in Reinforcement Learning for Reasoning
Wang Yang, Shouren Wang, Chaoda Song +6
Group Relative Policy Optimization (GRPO) has become a key technique for improving reasoning abilities in large language models, yet its behavior under different domain sequencing…
Mid-Think: Training-Free Intermediate-Budget Reasoning via Token-Level Triggers
Wang Yang, Debargha Ganguly, Xinpeng Li +5
Hybrid reasoning language models are commonly controlled through high-level Think/No-think instructions to regulate reasoning behavior, yet we found that such mode switching is lar…
Demystifying Hybrid Thinking: Can LLMs Truly Switch Between Think and No-Think?
Shouren Wang, Wang Yang, Xianxuan Long +4
Hybrid thinking enables LLMs to switch between reasoning and direct answering, offering a balance between efficiency and reasoning capability. Yet our experiments reveal that curre…
SELF: Self-Extend the Context Length With Logistic Growth Function
Phat Thanh Dang, Saahil Thoppay, Wang Yang +3
Large language models suffer issues when operated on long contexts that are larger than their training context length due to the standard position encoding for tokens in the attent…
Longer Context, Deeper Thinking: Uncovering the Role of Long-Context Ability in Reasoning
Wang Yang, Zirui Liu, Hongye Jin +3
Recent language models exhibit strong reasoning capabilities, yet the influence of long-context capacity on reasoning remains underexplored. In this work, we hypothesize that curre…
Speculative Thinking: Enhancing Small-Model Reasoning with Large Model Guidance at Inference Time
Wang Yang, Xiang Yue, Vipin Chaudhary +1
Recent advances leverage post-training to enhance model reasoning performance, which typically requires costly training pipelines and still suffers from inefficient, overly lengthy…
Thinking Preference Optimization
Wang Yang, Hongye Jin, Jingfeng Yang +2
Supervised Fine-Tuning (SFT) has been a go-to and effective method for enhancing long chain-of-thought (CoT) reasoning in relatively small LLMs by fine-tuning them with long CoT re…