© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 3842
    Prev
    Next
    INDEX
    Explanations

    usable / unusable / playable

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

    The neuron activates on words ending with the “-able”/“-ible” suffix (e.g., usable, potable, drinkable, unusable).

    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
     ſame
    -1.73
    haviours
    -1.63
     —
    -1.63
    跸
    -1.61
    мии
    -1.59
    TintMode
    -1.54
     bawat
    -1.52
    嚒
    -1.52
    邶
    -1.51
    ův
    -1.48
    POSITIVE LOGITS
     Those
    1.64
     that
    1.63
     eigentlichen
    1.52
     situation
    1.48
     ulter
    1.44
    "
    1.38
    Without
    1.36
    Only
    1.36
     However
    1.35
     barely
    1.34
    Activations Density 0.055%

    No Known Activations