Search papers, labs, and topics across Lattice.
Affiliation:
3
0
6
1
ImpossibleRubrics is introduced, a benchmark of 169 impossible tasks spanning six impossibility categories, each paired with a verifiable oracle certificate specifying what an honest answer may and may not claim, together with 48 answerable controls.
Query-driven extraction reveals that existing Vision-Language Models falter in multimodal tasks, but a collaborative framework can bridge the performance gap.
LLMs that ace general web browsing still fail miserably at autonomous scientific literature discovery, revealing a critical gap in research-oriented AI agent capabilities.
