© 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. 58676
    Prev
    Next
    INDEX
    Explanations

    said than done

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

    This neuron spots key words that occur in common multiword idiomatic or formulaic expressions (e.g. “no denying the fact,” “with this in mind,” “easier said than done”).

    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
     to
    -1.98
     be
    -1.62
     or
    -1.55
    Contexte
    -1.44
     інших
    -1.40
     our
    -1.39
    because
    -1.38
     pagare
    -1.36
    without
    -1.35
     diğer
    -1.34
    POSITIVE LOGITS
     groote
    1.36
    "../
    1.34
     erweitert
    1.32
    alucía
    1.31
    ساز
    1.30
    𞤮
    1.30
    \-
    1.27
    (+)
    1.23
     diretto
    1.23
    ...');
    1.23
    Activations Density 0.097%

    No Known Activations