© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 27507
    Prev
    Next
    INDEX
    Explanations

    tachyons or transistors

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

    This neuron detects the constituent pieces of long, domain-specific compound terms and proper names (e.g. multi-token technical or scientific words).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Kenapa
    -1.06
     docteur
    -0.93
     royaume
    -0.81
    gbaar
    -0.80
    Siapa
    -0.80
     turnips
    -0.80
    Shaft
    -0.80
    Reino
    -0.79
    гг
    -0.79
     père
    -0.78
    POSITIVE LOGITS
     and
    1.04
     instead
    0.92
     technique
    0.90
     edn
    0.88
     material
    0.87
     type
    0.86
    type
    0.84
     pocz
    0.82
    休み
    0.80
     breite
    0.80
    Activations Density 0.190%

    No Known Activations