© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 7680
    Prev
    Next
    INDEX
    Explanations

    really to emphasize action

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

    The neuron flags high‐frequency sentiment intensifiers—words like “really” or “easy” that emphasize or amplify a statement.

    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
    ه
    1.67
    i
    1.57
    ت
    1.31
    然後
    1.28
    ك
    1.27
    iD
    1.26
     BID
    1.23
    ič
    1.21
    BID
    1.21
    ように
    1.20
    POSITIVE LOGITS
    ﻭ
    1.52
    ﻙ
    1.30
    paar
    1.28
    ﺭ
    1.22
    testid
    1.20
    ﻥ
    1.20
    tested
    1.18
    lungen
    1.16
    нуться
    1.16
    ﺕ
    1.16
    Activations Density 0.065%

    No Known Activations