finding
active
finding:trackeragent-win-rate-53-6TrackerAgent win rate=53.6%
TrackerAgent won more than half its games
Source paper
extracted_from(2026) · Robert Müller · Clemens Müller
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- TrackerAgent won over half of the combined-comp1 games.
- Best code agent outperforming six of seven LLMs.
- TrackerAgent has the highest buy-right rate among all agents.
- In the 98-game slice, TrackerAgent had a higher win rate or TrueSkill than all LLMs except Gemini 3 Flash.
- Sonnet's win rate in exploratory games
- Calibration that conditional logic can beat cost-efficient LLMs in this setting.
- Code agents trade bargaining precision for acquisition pressure.
- deterministic code agent that maintains perfect information from observable events and makes greedy decisions conditioned on card counts and estimated wealth