© 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. 75884
    Prev
    Next
    INDEX
    Explanations

    carcinoma | hearing | sleep

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

    code-related patterns in programming files, especially whitespace and formatting elements.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron fires on code‐style identifiers (tokens like “dispatch_”, “get_”, or other programming names with underscores or mixed punctuation).

    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
    声音
    -0.73
     oh
    -0.71
     Models
    -0.71
     copy
    -0.69
    isata
    -0.67
     source
    -0.66
    皇子
    -0.65
     Ju
    -0.65
     میخ
    -0.65
     execution
    -0.65
    POSITIVE LOGITS
     Hearings
    0.78
     Nug
    0.78
    UMN
    0.71
     casters
    0.71
     cacti
    0.69
     Penalties
    0.69
     remastered
    0.68
    ABASE
    0.68
    onces
    0.67
     cyclists
    0.67
    Activations Density 0.066%

    No Known Activations