LongHarness Bench: Stress-Testing Language Model Harnesses for Long-Context Reasoning
Language-model (LM) harnesses enable LMs to operate effectively over long contexts using additional compute. However, existing long-context evaluations are insufficient for distinguishing modern harnesses, reflected by saturated accuracy across harnesses and largely similar evaluation costs. In this paper, we introduce...
Quang Hieu Pham, Thuy Duong Nguyen, Jocelyn Qiaochu Chen et al.
· 0 citations