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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 61480
    Prev
    Next
    INDEX
    Explanations

    something like this

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

    The neuron fires on vague placeholder words—indefinite pronouns and nouns like “something,” “anything,” “things,” “items,” etc., that refer to unspecified content.

    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
     فيلم
    -1.32
     "
    -1.32
     HttpHeaders
    -1.31
    тувати
    -1.28
    Through
    -1.26
     под
    -1.20
    ţin
    -1.20
     „
    -1.19
    施行
    -1.16
     не
    -1.15
    POSITIVE LOGITS
     that
    1.84
     it
    1.73
     from
    1.71
     such
    1.59
     this
    1.35
     andet
    1.29
     about
    1.23
     jaya
    1.23
     spalle
    1.21
     pietre
    1.19
    Activations Density 0.278%

    No Known Activations