conftrace_

output reliability

2 papers

Explore in graph

Co-occurring keywords · papers mentioning both

large language model (2) task-specific vulnerability (1) model vulnerability (1) semantic agreement (1) llm cascade (1) deferral mechanism (1) token-level confidence (1) template overfitting (1) instruction template (1) superficial expression (1)

Papers

Semantic Agreement Enables Efficient Open-Ended LLM Cascades EMNLP 2025 Superfluous Instruction: Vulnerabilities Stemming from Task-Specific Superficial Expressions in Instruction Templates ACL 2025