EyeMulator augments CodeLLM fine-tuning loss with token weights derived from human eye-tracking scan paths, producing large gains on code translation and summarization across StarCoder, Llama-3.2 and DeepSeek-Coder.
InProceedings of the 14th ACM / IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM)(Bari, Italy)(ESEM ’20)
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.SE 3verdicts
UNVERDICTED 3representative citing papers
A controlled eye-tracking study finds that code priority affects review time, cognitive load, and perceived quality but not reuse decisions, while author reputation changes visual attention patterns without altering performance or reuse choices.
Eye-tracking experiment finds that labeling code as LLM-generated increases fixation time without changing review thoroughness, with reviewers adapting criteria or using the prompt.
citing papers explorer
-
EyeMulator: Improving Code Language Models by Mimicking Human Visual Attention
EyeMulator augments CodeLLM fine-tuning loss with token weights derived from human eye-tracking scan paths, producing large gains on code translation and summarization across StarCoder, Llama-3.2 and DeepSeek-Coder.
-
An Eye for Trust: An Exploration of Developers' Trust Perceptions Through Urgency and Reputation
A controlled eye-tracking study finds that code priority affects review time, cognitive load, and perceived quality but not reuse decisions, while author reputation changes visual attention patterns without altering performance or reuse choices.
-
Same Scrutiny, More Time: Eye Tracking Insights into Reviewing LLM-Labelled Code
Eye-tracking experiment finds that labeling code as LLM-generated increases fixation time without changing review thoroughness, with reviewers adapting criteria or using the prompt.