© 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. 24-GEMMASCOPE-2-RES-16K
    4. 825
    Prev
    Next
    INDEX
    Explanations

    The neuron flags vague “placeholder” phrases built from a general quantifier (some, more, do, being) plus a common abstract noun or pronoun (stage, ways, information, so, held)—i.e. hedge-style referential expressions.

    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_24_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
    黢
    0.31
    サイド
    0.31
    殓
    0.31
    してきた
    0.30
     সাম
    0.30
     bauen
    0.30
     высокий
    0.28
     बेसिकली
    0.28
    ować
    0.28
     हैंड
    0.28
    POSITIVE LOGITS
     genom
    0.33
     objectForKey
    0.32
    Е
    0.32
    ሔ
    0.32
    מה
    0.32
    이가
    0.32
     ।।
    0.32
    니다
    0.31
    ться
    0.31
    เร็จ
    0.31
    Activations Density 0.000%

    No Known Activations