⬢github first seen 4 d ago, last 2 d ago, peak #6
Open-source tool strips censorship and sycophancy from language models
Original: p-e-w/heretic
A developer publishing under the name p-e-w has released Heretic, a Python tool that modifies language models without additional training. It claims to remove censorship, suppress low-quality repetitive output described as slop, and reduce sycophantic responses. The project has drawn attention in developer communities, with users debating the trade-offs between model compliance and safety guardrails.
Why now: Debate over AI censorship and model behaviour is currently a hot topic among developers experimenting with open-weight models
Rank over time, top of the chart is #1. 196 snapshots from 4 d ago to 2 d ago.
Evidence
- p-e-w/heretic · p-e-w · 16
Training-free language model modification: Remove censorship, suppress slop, reduce sycophancy, and more
API: https://socialmediatrends-api.osmike.com/v1/trends/4610472