© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 2339
    Prev
    Next
    INDEX
    Explanations

    indicates probability or assumption

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

    This neuron detects hedging or speculative language—words that signal uncertainty or conjecture (e.g. evidently, perhaps, must have had some difficulty).

    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
    ൈ
    0.64
    تان
    0.62
    り
    0.61
    なんと
    0.58
     credito
    0.58
    idual
    0.58
    Validity
    0.58
    cc
    0.57
    ги
    0.57
    ო
    0.57
    POSITIVE LOGITS
     mindestens
    0.70
     jenis
    0.68
     mellom
    0.67
     mellem
    0.67
     mellan
    0.66
     någon
    0.66
     tämä
    0.64
     nogen
    0.64
     koska
    0.64
     destiné
    0.64
    Activations Density 0.075%

    No Known Activations