Skip to content
View SolomonB14D3's full-sized avatar
2 followers · 4 following

Block or report SolomonB14D3

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
SolomonB14D3/README.md

Bryan Sanchez — Independent ML Researcher

Building tools that find where AI models are wrong and fix them.

Papers

  • Correcting Suppressed Log-Probabilities in Language Models with Post-Transformer Adapters (2026) A 786K-parameter post-transformer adapter corrects censorship-suppressed log-probabilities across Qwen3-4B, 8B, and 14B. Includes documentation of a silent gradient bug in Apple MLX. Code · Weights

Repos

Repo What it does
qwen-adapter-correction Post-transformer adapters for correcting suppressed log-probabilities in language models.
noethersolve 300+ verified computational tools for AI agents via MCP. PyPI.
knowledge-fidelity Behavioral auditing toolkit for LLMs. PyPI.

Pinned Loading

  1. knowledge-fidelity knowledge-fidelity Public

    Behavioral auditing & repair toolkit for LLMs. Measures 8 dimensions via confidence probes.

    Python 3