© 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. 31-GEMMASCOPE-2-RES-16K
    7. 6771
    Prev
    Next
    INDEX
    Explanations

    reveal or clue

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

    The neuron fires on verbs and nouns that denote acts of uncovering or revealing information (e.g. reveal, expose, dig, dissect).

    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
    Hochspringen
    0.54
    rax
    0.53
    と感じ
    0.53
    acement
    0.53
     accented
    0.52
    ίκη
    0.52
     paik
    0.50
     센
    0.49
    认可
    0.49
    amped
    0.48
    POSITIVE LOGITS
     reveal
    2.69
     revealing
    2.58
    reveal
    2.50
     Reveal
    2.42
     revelar
    2.42
    Reveal
    2.38
     revelations
    2.36
     reveals
    2.33
     revealed
    2.31
     Reveals
    2.28
    Activations Density 0.628%

    No Known Activations