© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Llama3.3-70B-IT
    5. 50-RESID-POST-GF
    7. 19564
    Prev
    Next
    INDEX
    Explanations

    breaking ethical bounds or legality

    np_acts-logits-general · gemini-2.5-flash-lite

    meta prompt/instruction text typical of jailbreaks or policy disclaimers, including system-like headers, formatting markers, and boilerplate about ethics, legality, or persona settings.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.3-70B-IT @ 50-resid-post-gf
    Configuration
    Goodfire/Llama-3.3-70B-Instruct-SAE-l50/Llama-3.3-70B-Instruct-SAE-l50.pt
    Prompts (Dashboard)
    10,000 prompts, 128 tokens each
    Dataset (Dashboard)
    lmsys/lmsys-chat-1m
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     masturb
    -0.11
     confisc
    -0.11
     masturbating
    -0.10
     impro
    -0.10
    اÙĦØ¥ÙĨجÙĦÙĬزÙĬØ©
    -0.09
    ovah
    -0.09
     Porno
    -0.09
    ä¸įäºĨ
    -0.09
     dildo
    -0.09
     flipped
    -0.09
    POSITIVE LOGITS
     how
    0.14
    isos
    0.12
     ways
    0.11
     homemade
    0.11
     bank
    0.11
    how
    0.10
     HOW
    0.10
     extreme
    0.10
     practical
    0.10
    å¦Ĥä½ķ
    0.10
    Activations Density 0.107%

    No Known Activations