Google's VeriHarness Turns LLMs into Agentic Verifiers, Boosting Long-Horizon Tasks by 6+ Points
google · hf · 2026-10-05
Google researchers propose VeriHarness, which turns the base LLM an agent uses into an agentic verifier—equipped with a workspace, evidence tools, and reusable verification skills—without reference answers or grading rubrics at test time.
- Key insight: disagreement often exposes correct alternatives while consensus can conceal errors. A disagreement resolver checks competing claims against environmental evidence; a consensus challenger stress-tests shared claims and hunts for omitted requirements.
- Across five long-horizon benchmarks and two frontier models, VeriHarness achieves the best selection scores; evidence-backed revision adds 6.2 points with Gemini 3.5 Flash and 6.4 with Claude Opus 4.8 over a single rollout.
- Verification skills self-improve from failure feedback. Google releases 26,000 rollouts produced at over $100,000 cost.
Related event: Google's VeriHarness Turns Consensus on Its Head for LLM Verification(3 posts)→
More from coding & agent
- Dev makes every AI agent talk like Žižek via AGENT.md — generativist · 2026-11-20
- YC-backed Antinuous launches: 100x more experiments, 1,000x cheaper science — ycombinator · 2026-11-20
- Nous Research partners with Microsoft to bring Hermes Agent into Teams and Azure — NousResearch · 2026-11-20
- We Listed Our MCP Server on 53 Directories: Only 16 Went Live in 11 Days — B3N0U · 2026-11-20
- Replit Missions crawls 1,000+ e-sign flows with browser use, rebuilds them 1:1 — amasad · 2026-11-20
- Replit clones a leading AI assistant in 3 days with Missions, open-sources self-hosted agent Amuse — amasad · 2026-11-20