© 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-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 1701
    Prev
    Next
    INDEX
    Explanations

    why did

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

    The neuron isn’t detecting any specific semantic category but rather fires strongly on very common function and high-frequency words (e.g. short pronouns, prepositions, auxiliary verbs, and other “glue” words like “with,” “they,” “are,” “find,” “go,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     corresponden
    0.57
    が必要です
    0.56
     mouthful
    0.56
    !";
    0.54
     রায়
    0.54
     отсутствует
    0.53
     事
    0.52
     trouxe
    0.52
    太过
    0.52
    太
    0.52
    POSITIVE LOGITS
     their
    0.62
     THEIR
    0.59
     نها
    0.55
     leur
    0.54
    férences
    0.53
    WithOwner
    0.52
    rac
    0.50
    εί
    0.49
    Their
    0.49
    their
    0.49
    Activations Density 0.027%

    No Known Activations