evaluation-package · Source-linked discovery
FRAMES — Baseline
FRAMES evaluates multi-hop reasoning by requiring models to answer questions that can only be resolved by chaining information across multiple Wikipedia documents. In the baseline variant, all source documents are provided upfront, testing whether a model can reason across a full document set to produce the correct answer. Scored on answer accuracy.
OriginSatyapriya Krishna, Kalpesh Krishna, Anhad Mohananey et al.TopicsGeneral capabilityStatuscatalogued
Can support
Not independently assessed by FronteraEval yet.
Cannot support by itself
No inference beyond the upstream source should be made until the protocol is reviewed.