© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 43918
    Prev
    Next
    INDEX
    Explanations

    identifying nouns followed by descriptive verbs

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

    The neuron fires strongly on key technical or document‐structure nouns (like “condition,” “file,” “invention,” “circuit,” “sensor,” etc.) that typically label sections or objects in formal/specification texts.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     of
    -2.33
     this
    -2.14
     with
    -1.72
     between
    -1.70
    際には
    -1.63
     within
    -1.57
    鵟
    -1.53
     these
    -1.52
     that
    -1.52
     on
    -1.52
    POSITIVE LOGITS
    <bos>
    1.68
    
    1.63
     них
    1.63
     depois
    1.63
    最近
    1.61
    ocardium
    1.60
    ()],
    1.58
     trasera
    1.57
    昨日
    1.55
    }
    
    1.54
    Activations Density 0.658%

    No Known Activations