luulinh90s 1 karma 6d on HN HN profile →
Coverage
We've seen 1 of an unknown number of submissions
Full eval: 1 Lite-only: 0 Unevaluated: 0
Deep Read section-level analysis · 1 full evaluation
HRCB +0.35
E-full / S +0.29 / +0.52 SETL +0.02
UDHR Fingerprint Preamble: +0.35 Article 1: +0.15 Article 2: ND Article 3: ND Article 4: ND Article 5: ND Article 6: +0.10 Article 7: ND Article 8: ND Article 9: ND Article 10: ND Article 11: ND Article 12: +0.05 Article 13: ND Article 14: ND Article 15: ND Article 16: ND Article 17: ND Article 18: +0.35 Article 19: +0.59 Article 20: ND Article 21: ND Article 22: ND Article 23: +0.10 Article 24: ND Article 25: ND Article 26: +0.38 Article 27: +0.58 Article 28: +0.20 Article 29: +0.35 Article 30: ND
1 stories
1.
HRCB +0.35
E +0.29
S +0.52
Steering interpretable language models with concept algebra (www.guidelabs.ai)
75 points by luulinh90s 4 days ago | 8 comments | hrcb v3.7 AI Transparency & Scientific Access