© 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. 12781
    Prev
    Next
    INDEX
    Explanations

    coming to come

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

    The neuron fires on discourse‐level transition phrases—especially summary/conclusion markers like “In the end,” (and similar transitional cues).

    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
     maxit
    0.60
    래요
    0.54
     Luke
    0.54
    一样
    0.53
     poni
    0.53
     Lorenz
    0.53
     qualifiers
    0.52
    珢
    0.52
     Lessee
    0.51
    麼
    0.51
    POSITIVE LOGITS
    Exerc
    0.63
    :.
    0.55
    DEPENDENCIA
    0.55
    fetchAll
    0.52
    하지만
    0.52
     bring
    0.51
    adas
    0.51
     samtid
    0.51
     Kategor
    0.51
    jeux
    0.51
    Activations Density 0.000%

    No Known Activations