Mycelium meta-study reframes hallucination detection with new Span Coverage F1 metric
Julia Belikova · hf · 2026-10-27
A meta-study from Skoltech and Sber AI Lab treats hallucination detection as a benchmark-conditioned measurement problem, organizing existing methods along provenance, typology, and granularity axes. The paper proposes Span Coverage F1, a new metric designed to fix the variability of span-level evaluation and enable fairer cross-method comparison.
More from Research
- Gowers: OpenAI's math drop includes Kakeya proof; AGMAI urges release of Yang-Mills outputs — calabi_and_yau · 2026-10-27
- Protein LM embeddings + similarity-based label augmentation improve pKa prediction — bravo_abad · 2026-10-27
- New eLife study makes whole-brain mean-field models more biologically realistic — AnnaCiaunica · 2026-10-27
- GenXR-Bench: AI coding agents fail at XR apps, best model scores just 50.9% — rohanpaul_ai · 2026-10-27
- EmbodiedLab Benchmarks Robots That Must Act to Know in Simulated Labs — HexaFuture · 2026-10-27
- Zhejiang University's RoboHarness Lets LLM Agents Beat VLA Models on BEHAVIOR — zju · 2026-10-27