© 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. 7304
    Prev
    Next
    INDEX
    Explanations

    Arch Linux package installation

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

    This neuron fires most strongly on unusually rare or infrequent tokens (i.e. it highlights low-frequency, “surprising” words).

    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
    </em>
    0.71
     Schein
    0.70
     robust
    0.70
     Vend
    0.70
    я
    0.66
     frontline
    0.66
    βε
    0.66
    -
    0.66
    くちゃ
    0.65
     shortage
    0.65
    POSITIVE LOGITS
     indiqu
    0.91
     agradable
    0.91
    toluene
    0.91
     apparaissent
    0.87
     invertebr
    0.86
    GraphQL
    0.84
    𒆳
    0.84
    satz
    0.83
     derivada
    0.82
     explique
    0.82
    Activations Density 0.000%

    No Known Activations