© 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. 34-GEMMASCOPE-RES-131K
    7. 31954
    Prev
    Next
    INDEX
    Explanations

    specific words following certain terms

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

    The neuron fires strongly on non‐natural‐language tokens—particularly digits, underscores, and other code‐ or metadata‐style identifiers (e.g. version numbers, IDs, or file/path markers).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     anzunehmen
    -0.75
    砼
    -0.74
     Peterborough
    -0.68
     dilemma
    -0.66
     bites
    -0.65
    puppy
    -0.65
    INARY
    -0.65
     + 
    -0.64
    אַ
    -0.64
     Hector
    -0.64
    POSITIVE LOGITS
    zeption
    0.75
    症状
    0.74
     الجزائر
    0.73
     Melville
    0.73
     Wadsworth
    0.73
     Effort
    0.70
    jā
    0.70
     deposit
    0.69
     effort
    0.69
     perf
    0.69
    Activations Density 0.010%

    No Known Activations