Top New Best Ask Show Jobs

JakubCwi

5 karma · joined January 20, 2026

submissions comments

Why LLMs Lie in Different Ways?

orsonai.com·1 pts·JakubCwi·

Comparing confabulation signals across 8 LLMs

orsonai.com·2 pts·JakubCwi·

5-shot in-context examples are enough to blind your LLM

researchgate.net·1 pts·JakubCwi·

Why LLMs stop learning from examples in long contexts: The ICL Collapse

orsonai.com·3 pts·JakubCwi·

Paper: Detecting hallucinations before the first token

researchgate.net·4 pts·JakubCwi·

AI models know they are guessing before they generate text

orsonai.com·4 pts·JakubCwi·

The containment box that destroys the AI if it tries to think its way out

redact-app.com·3 pts·JakubCwi·

Hashgraph Detection Key – If you touch it, it breaks

redact-app.com·2 pts·JakubCwi·

AI box that fights back

researchgate.net·2 pts·JakubCwi·

10⁻²³ hurts AI

redact-app.com·1 pts·JakubCwi·