Search papers, labs, and topics across Lattice.
3
0
7
3
LLMs exhibit a pervasive optimism bias when evaluating research proposals, frequently rating methodologically unsound ideas as promising, suggesting they're not ready to replace human reviewers.
Evaluative language in LLM-generated resume summaries can introduce significant biases that traditional fairness audits might miss, destabilizing hiring processes.
Instead of imitating reflections, LLM agents can be trained to reason about action quality by rewarding correct judgments between alternative actions, leading to improved performance and generalization.
