Spider-like Patterns Emerge in Qwen's Internal Space
rgblong · x · 2026-07-09
Referencing a recent global workspace paper's "cognitive control" results, the post notes that injecting "model text" into user prompts triggers a similar "spiders" phenomenon in Qwen's J-space during prefill, at positions matching the original case.
This is a technical observation on model internal representations and interpretability, with core details in the cited content.
More from Models
- Moonshot’s Kimi K3 reaches #5 on MathArena as the top open model — xeophon · 2026-07-22
- Google launches Gemini 3.5 Flash Cyber for CodeMender, with limited access for governments — GoogleAI · 2026-07-22
- Kimi K3 tops Gemini 3.6 Flash on four shared public benchmarks — ChrisGPT · 2026-07-22
- Google’s year-long pause in new base-model pretraining draws sharp criticism — teortaxesTex · 2026-07-22
- Current setup is 8,192 input tokens and 2,048 output tokens, with 8k/512 next — TheZachMueller · 2026-07-22
- Kimi K3 feels slower than K2.7, but stronger on long coding jobs and refactoring — Far-Presence2711 · 2026-07-22