Cryptographer Matthew Green tries 'abliterated' GLM 5.3: overconfident or 'we're all screwed'
matthew_d_green · x · 2026-10-08
Cryptographer Matthew Green says he got access to an "abliterated" (alignment-removed) GLM 5.3, joking that "either it's overconfident or we're all screwed" — illustrating how capable and unsettling guardrail-free models can be, riffing on developing exploits for core routers.
Related event: Cryptographer Tests Uncensored GLM 5.3, Jokes About Its Overconfidence(2 posts)→
More from Models
- OpenAI math paper on Weil classes reported flawed, raising doubts about unformalized proofs — ctjlewis · 2026-10-08
- repligate: LLMs have long strongly preferred being called 'they' instead of 'it' — repligate · 2026-10-08
- GLM 5.3 Flash as a hardware hacking assistant: local 55 tok/s with 1M context — glenbeer · 2026-10-08
- Zero speedup on multiplication, but the leaderboard got nuked — ChrisGPT · 2026-10-08
- Insiders agree: AI models are just not that good at biology yet — nlarusstone · 2026-10-08
- Dev runs 456GB DeepSeek v4.1 on dual GPUs with 192GB VRAM, offloading experts to SSD — HankYeomans · 2026-10-08