© 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. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 111303
    Prev
    Next
    INDEX
    Explanations

    rat and rat-

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

    The neuron activates for variations of the word "rat" including "Rat", "rat", "ratchet", "ratfish", and names like "Ratty" or "Ratner".

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    𖤣
    -2.98
    櫚
    -2.75
    jų
    -2.72
    ztes
    -2.69
    vantaged
    -2.66
     моя
    -2.59
    𓇼
    -2.55
    嫫
    -2.48
    龵
    -2.48
    話は
    -2.44
    POSITIVE LOGITS
    </h1>
    2.72
    }
    2.53
    </th>
    2.45
    ‐
    2.41
    {
    2.25
    gefügt
    2.20
    й
    2.19
    可以
    2.19
    '
    2.19
    牁
    2.16
    Activations Density 0.023%

    No Known Activations