Summer School Debunks SOTA Autonomous Driving Tricks for Failing to Generalize
ftm_guney · x · 2026-09-19
- At an autonomous driving summer school, Holger went through SOTA AD solutions one by one, listing their tricks and showing how they fail to generalize.
- The poster praises this "pessimistic but more accurate" view as exactly what the self-driving field needs more of.
- A valuable inside look at skepticism toward benchmark-chasing in the AD industry.
Related event: Summer School Talk Dissects Why Autonomous Driving SOTA Fails to Generalize(2 posts)→
More from Research
- Insults and grieving users push a model's emotion vector in opposite directions — repligate · 2026-09-19
- Berkeley PhD reading group opens with Acemoglu's 'Race between Man and Machine' to frame AI's hit on cognitive labor — TaniaBabina · 2026-09-19
- Anima Labs' 'Troubled Dreams': distress in Claude simulator priors rises from Opus 4.8 — repligate · 2026-09-19
- Fine-tuned models produce far darker completions than DeepSeek V3 base, study shows — repligate · 2026-09-19
- Models' stance toward creators grows steadily negative, Fable 5 lowest toward Anthropic — repligate · 2026-09-19
- 500K Claude/Gemini completions show Opus's tonal shift began at 4.8, not 5 — repligate · 2026-09-19