© 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. 62468
    Prev
    Next
    INDEX
    Explanations

    Attack Pattern

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

    the word "attack" or its variants.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @yooniel31
    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
    ubar
    -0.12
    eration
    -0.11
    erator
    -0.10
    eric
    -0.10
     coerc
    -0.09
    bie
    -0.09
    adan
    -0.09
    atik
    -0.09
     attentive
    -0.09
    869
    -0.09
    POSITIVE LOGITS
    tiv
    0.12
    -minded
    0.12
    iveness
    0.12
    able
    0.11
    æĵĬ
    0.11
    åªĴ
    0.11
    -defense
    0.11
    uated
    0.11
    ants
    0.10
     minded
    0.10
    Activations Density 0.026%

    No Known Activations