© 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. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 73209
    Prev
    Next
    INDEX
    Explanations

    quantifiable results

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

    The neuron fires on quantitative metrics—numbers, statistics, and words reporting results or measurements.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     but
    -1.64
    announced
    -1.40
    但是
    -1.38
     huile
    -1.34
    ländische
    -1.24
     mariposas
    -1.23
     paille
    -1.20
     doInBackground
    -1.20
     odeur
    -1.20
     mutta
    -1.19
    POSITIVE LOGITS
     our
    1.82
     last
    1.52
     Gewebe
    1.37
     over
    1.28
    itored
    1.26
     Tbsp
    1.25
     percent
    1.20
     overall
    1.16
     across
    1.16
     Stoffe
    1.16
    Activations Density 0.031%

    No Known Activations