© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Llama3.3-70B-IT
    5. 50-RESID-POST-GF
    7. 25074
    Prev
    Next
    INDEX
    Explanations

    names and titles

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

    stems of content words—especially nouns, proper names, or template field labels—often appearing at the beginning of a word or token.

    oai_token-act-pair · gpt-5Triggered by @bestetascii

    The neuron identifies specific lexical items, often proper nouns, technical terms, or distinct non-English words.

    oai_token-act-pair · gemini-2.5-flashTriggered by @bestetascii
    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
    ¥IJ
    -0.08
    ξÏį
    -0.08
     müc
    -0.07
    Ã¶ÄŁ
    -0.07
    UTURE
    -0.07
    =}
    -0.07
    üç
    -0.07
    /grpc
    -0.07
    опаÑģ
    -0.07
    uibModal
    -0.07
    POSITIVE LOGITS
    ses
    0.08
    /of
    0.08
    Ìģ
    0.07
    еÑĢим
    0.07
    ifies
    0.07
    o
    0.07
    olson
    0.07
    =”;
    0.07
    ££
    0.07
    -all
    0.07
    Activations Density 0.533%

    No Known Activations