© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 94033
    Prev
    Next
    INDEX
    Explanations

    racist, racial, racism, homophobic

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

    The neuron strongly activates on numeric tokens (digits, counts, years) rather than ordinary 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_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
    Asser
    -0.83
     законо
    -0.78
     tristeza
    -0.74
    anggung
    -0.72
    蛯
    -0.71
    デメリット
    -0.69
     might
    -0.69
     Flach
    -0.69
    Renewable
    -0.68
    風險
    -0.68
    POSITIVE LOGITS
     racist
    2.39
     racial
    2.28
     racially
    1.88
     racism
    1.84
    racial
    1.76
     Racial
    1.70
    racist
    1.52
    Rac
    1.48
     homophobic
    1.42
     Racism
    1.41
    Activations Density 0.029%

    No Known Activations