3 citations · 3 across the 8 of their papers we have counts for
11 papers
Calibrated Typed Decisions for Entity, Relation, and Event Extraction
Yang Wang, Hanlin Xiao, Yanan Ma +2
Information extraction models often perform well on the datasets they are trained on, but transferring to new domains and annotation guidelines remains challenging. Supporting user…
Rate-Resolved Text Evaluation with Masked Diffusion Language Models
Yiqi Liu, Kun Zhao, Chenghua Lin +1
A single evaluation score can rank generation systems but rarely reveals what failed. We propose rate-resolved text evaluation, which turns the masking rate of masked diffusion lan…
Feedback as an Interface: Controlling Form and Detail in LLM Translation Refinement
Yiqi Liu, Flavjo Tomori, Nafise Sadat Moosavi +2
Feedback presentation shapes how LLMs use evaluator judgments during translation refinement. We introduce a machine translation protocol that keeps the evaluator score, main issue,…
When Tool-Backed Skill Retrieval Fails: Source-Style Collapse in Executable Capability Retrieval
Yiqi Liu, Joseph James, Yang Wang +2
Large-scale agents increasingly rely on retrieval to access external capabilities. We study this retrieval gate in structured tools and APIs, a measurable class of tool-backed exec…
Broken Symmetry in LLM Refusal: Answer Release Is More Local Than Refusal Restoration
Yiqi Liu, Yang Wang, Songxin Wang +2
When a language model refuses to answer a prompt, it is unclear whether the correct answer is erased from its internal representations, or merely suppressed at the output layer. We…
Can LLMs Write Reliable Rubrics? A Meta-Evaluation for Experiment Reproduction
Hanhua Hong, Yizhi Li, Jiaoyan Chen +4
Rubric-based evaluation is a promising approach for assessing open-ended outputs from LLM-based research agents, particularly in paper reproduction, where direct paper-to-repositor…
The Achilles' Heel of Angular Margins: A Chebyshev Polynomial Fix for Speaker Verification
Yang Wang, Yiqi Liu, Chenghao Xiao +1
Angular margin losses, such as AAM-Softmax, have become the de facto in speaker and face verification. Their success hinges on directly manipulating the angle between features and…
RIGOURATE: Quantifying Scientific Exaggeration with Evidence-Aligned Claim Evaluation
Joseph James, Chenghao Xiao, Yucheng Li +2
Scientific rigour tends to be sidelined in favour of bold statements, leading authors to overstate claims beyond what their results support. We present RIGOURATE, a two-stage multi…
Beyond One-Size-Fits-All: Inversion Learning for Highly Effective NLG Evaluation Prompts
Hanhua Hong, Chenghao Xiao, Yang Wang +3
Evaluating natural language generation systems is challenging due to the diversity of valid outputs. While human evaluation is the gold standard, it suffers from inconsistencies, l…
ContrastScore: Towards Higher Quality, Less Biased, More Efficient Evaluation Metrics with Contrastive Evaluation
Xiao Wang, Daniil Larionov, Siwei Wu +4
Evaluating the quality of generated text automatically remains a significant challenge. Conventional reference-based metrics have been shown to exhibit relatively weak correlation…
LLMs as Narcissistic Evaluators: When Ego Inflates Evaluation Scores
Yiqi Liu, Nafise Sadat Moosavi, Chenghua Lin
Automatic evaluation of generated textual content presents an ongoing challenge within the field of NLP. Given the impressive capabilities of modern language models (LMs) across di…