6 papers · 1 filter
When to Think, When to Speak: Learning Disclosure Policies for LLM Reasoning
Jiaqi Wei, Xuehang Guo, Pengfei Yu +5
In single-stream autoregressive interfaces, the same tokens both update the model state and constitute an irreversible public commitment. This coupling creates a silence tax: addit…
Reflection Pretraining Enables Token-Level Self-Correction in Biological Sequence Models
Xiang Zhang, Jiaqi Wei, Yuejin Yang +8
Chain-of-Thought (CoT) prompting has significantly advanced task-solving capabilities in natural language processing with large language models. Unlike standard prompting, CoT enco…
Unifying Tree Search Algorithm and Reward Design for LLM Reasoning: A Survey
Jiaqi Wei, Xiang Zhang, Yuejin Yang +10
Deliberative tree search is a cornerstone of modern Large Language Model (LLM) research, driving the pivot from brute-force scaling toward algorithmic efficiency. This single parad…
Why Prompt Design Matters and Works: A Complexity Analysis of Prompt Search Space in LLMs
Xiang Zhang, Juntai Cao, Jiaqi Wei +2
Despite the remarkable successes of large language models (LLMs), the underlying Transformer architecture has inherent limitations in handling complex reasoning tasks. Chain-of-tho…
Tokenization Constraints in LLMs: A Study of Symbolic and Arithmetic Reasoning Limits
Xiang Zhang, Juntai Cao, Jiaqi Wei +2
Tokenization is the first - and often underappreciated - layer of computation in language models. While Chain-of-Thought (CoT) prompting enables transformer models to approximate r…
Counting Ability of Large Language Models and Impact of Tokenization
Xiang Zhang, Juntai Cao, Chenyu You
Transformers, the backbone of modern large language models (LLMs), face inherent architectural limitations that impede their reasoning capabilities. Unlike recurrent networks, Tran…