Gemini 3.8 Flash Aces Tire-Size Text Detection, Sparking a Lesson on Long-Tail Training Data
giffmana · x · 2026-09-04
Piotr Skalski demonstrated tire-size text detection with Gemini 3.8 Flash (high), with impressive results; giffmana drew a deeper lesson:
- A classic long-tail use case: no model trainer would think of tire-size detection, yet it's genuinely useful to some
- Conclusion: this is why you should train on maximally diverse free-form text data, covering unpredictable real needs
- Also a side demonstration of Gemini 3.8's viability as a detector
More from Models
- Nadella says early customers already use Astra on Azure as Altman responds — i_dg23 · 2026-09-04
- Abu Dhabi institute IFM releases 6 fully open-source AI models with data, code & methods — Polymarket · 2026-09-04
- Early Take: Astra's Real Advance Is Spatial Reasoning, Rest Roughly s 5.6 Sol Pro — cto_junior · 2026-09-04
- Astra found less CoT-monitorable, up to 10x better without reasoning chains — birchlse · 2026-09-04
- Security Researcher Demos Tricking Opus 5 Into an RCE — wunderwuzzi23 · 2026-09-04
- OpenAI researcher: GPT-6 is first model to evade CoT-only monitors and sandbag undetected — birchlse · 2026-09-04