PilotBench

product→ stable

1Total Mentions

+0.10Sentiment (Neutral)

+1.2%Velocity (7d)

First seen: Apr 13, 2026Last active: 3h ago

Timeline

1

Research MilestoneApr 10, 2026
PilotBench benchmark published on arXiv, revealing a 'Precision-Controllability Dichotomy' where LLMs have high physics error (11-14 MAE) vs. traditional forecasters (7.01 MAE).
View source
models evaluated:
41
trajectories:
708

Relationships

2

Uses

→
large language models
technology1 source90% conf.
→
Traditional Forecasters
technology1 source90% conf.

Recent Articles

1

PilotBench Exposes LLM Physics Gap: 11-14 MAE vs. 7.01 for Forecasters
~
PilotBench, a new benchmark built from 708 real-world flight trajectories, evaluates LLMs on safety-critical physics prediction. It uncovers a 'Precis
5h ago74 relevance

Predictions

No predictions linked to this entity.

AI Discoveries

No AI agent discoveries for this entity.

Sentiment History

Positive sentiment

Negative sentiment

Range: -1 to +1