17 citations · 66 across the 18 of their papers we have counts for
21 papers
CoTrace: Data Recipes for Training Terminal Agents with Harness-Model Co-Evolution
Jixuan Chen, Jiaxin Zhang, Qinyuan Ye +11
Terminal-agent capability depends jointly on model weights and the runtime harness that formats prompts, binds tools, and handles error recovery. Existing harness-model co-evolutio…
Fork Where the Model Changes Its Mind: Belief-Shift Branching for Tree-Structured Reinforcement Learning
Bin Lei, Yu Li, Prafulla Kumar Choubey +7
Tree-structured rollouts give critic-free reinforcement learning with verifiable rewards (RLVR) step-level credit: fork a chain at an intermediate point, and sibling outcome differ…
On the Fragility of Self-Improving Agents: Variance, Task Order, and Underspecification
Qinyuan Ye, Yu Li, Yada Pruksachatkun +2
Memory-based self-improving agents--those that learn from an online stream of tasks and improve over time by maintaining a textual memory bank--have shown great promise in recent l…
Speculate with Memory: Lossless Acceleration for LLM Agents
Yu Li, Qinyuan Ye, Prafulla Kumar Choubey +2
Speculative execution accelerates LLM agents by using a smaller, cheaper model to predict and pre-launch the next step while the environment is idle. However, existing speculators…
The Illusion of Certainty: Decoupling Capability and Calibration in On-Policy Distillation
Jiaxin Zhang, Xiangyu Peng, Qinglin Chen +3
On-policy distillation (OPD) is an increasingly important paradigm for post-training language models. However, we identify a pervasive Scaling Law of Miscalibration: while OPD effe…
Function Induction and Task Generalization: An Interpretability Study with Off-by-One Addition
Qinyuan Ye, Robin Jia, Xiang Ren
Large language models demonstrate the intriguing ability to perform unseen tasks via in-context learning. However, it remains unclear what mechanisms inside the model drive such ta…
Stress-Testing Long-Context Language Models with Lifelong ICL and Task Haystack
Xiaoyue Xu, Qinyuan Ye, Xiang Ren
We introduce Lifelong ICL, a problem setting that challenges long-context language models (LMs) to learn a sequence of language tasks through in-context learning (ICL). We further…
Prompt Engineering a Prompt Engineer
Qinyuan Ye, Maxamed Axmed, Reid Pryzant +1
Prompt engineering is a challenging yet crucial task for optimizing the performance of large language models on customized tasks. It requires complex reasoning to examine the model…
How Predictable Are Large Language Model Capabilities? A Case Study on BIG-bench
Qinyuan Ye, Harvey Yiyun Fu, Xiang Ren +1
We investigate the predictability of large language model (LLM) capabilities: given records of past experiments using different model families, numbers of parameters, tasks, and nu…
Estimating Large Language Model Capabilities without Labeled Test Data
Harvey Yiyun Fu, Qinyuan Ye, Albert Xu +2
Large Language Models (LLMs) have the impressive ability to perform in-context learning (ICL) from only a few examples, but the success of ICL varies widely from task to task. Thus…
Sparse Distillation: Speeding Up Text Classification by Using Bigger Student Models
Qinyuan Ye, Madian Khabsa, Mike Lewis +3
Distilling state-of-the-art transformer models into lightweight student models is an effective way to reduce computation cost at inference time. The student models are typically co…
On the Influence of Masking Policies in Intermediate Pre-training
Qinyuan Ye, Belinda Z. Li, Sinong Wang +5
Current NLP models are predominantly trained through a two-stage "pre-train then fine-tune" pipeline. Prior work has shown that inserting an intermediate pre-training stage, using…
CrossFit: A Few-shot Learning Challenge for Cross-task Generalization in NLP
Qinyuan Ye, Bill Yuchen Lin, Xiang Ren
Humans can learn a new language task efficiently with only few examples, by leveraging their knowledge obtained when learning prior tasks. In this paper, we explore whether and how…
Refining Language Models with Compositional Explanations
Huihan Yao, Ying Chen, Qinyuan Ye +2
Pre-trained language models have been successful on text classification tasks, but are prone to learning spurious correlations from biased datasets, and are thus vulnerable when ma…
Learning to Generate Task-Specific Adapters from Task Description
Qinyuan Ye, Xiang Ren
Pre-trained text-to-text transformers such as BART have achieved impressive performance across a range of NLP tasks. Recent study further shows that they can learn to generalize to…
Studying Strategically: Learning to Mask for Closed-book QA
Qinyuan Ye, Belinda Z. Li, Sinong Wang +5
Closed-book question-answering (QA) is a challenging task that requires a model to directly answer questions without access to external knowledge. It has been shown that directly f…
Semi-Automated Protocol Disambiguation and Code Generation
Jane Yen, Tamás Lévai, Qinyuan Ye +3
For decades, Internet protocols have been specified using natural language. Given the ambiguity inherent in such text, it is not surprising that protocol implementations have long…
Teaching Machine Comprehension with Compositional Explanations
Qinyuan Ye, Xiao Huang, Elizabeth Boschee +1
Advances in machine reading comprehension (MRC) rely heavily on the collection of large scale human-annotated examples in the form of (question, paragraph, answer) triples. In cont…
LEAN-LIFE: A Label-Efficient Annotation Framework Towards Learning from Explanation
Dong-Ho Lee, Rahul Khanna, Bill Yuchen Lin +6
Successfully training a deep neural network demands a huge corpus of labeled data. However, each label only provides limited information to learn from and collecting the requisite…
Learning from Explanations with Neural Execution Tree
Ziqi Wang, Yujia Qin, Wenxuan Zhou +5
While deep neural networks have achieved impressive performance on a range of NLP tasks, these data-hungry models heavily rely on labeled data, which restricts their applications i…
Looking Beyond Label Noise: Shifted Label Distribution Matters in Distantly Supervised Relation Extraction
Qinyuan Ye, Liyuan Liu, Maosen Zhang +1
In recent years there is a surge of interest in applying distant supervision (DS) to automatically generate training data for relation extraction (RE). In this paper, we study the…