© 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. 50423
    Prev
    Next
    INDEX
    Explanations

    typical sequences

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

    The neuron is responding to code‐style or programming syntax, in particular parameter assignments and technical indicator expressions (e.g. “time_period=5”).

    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
     announce
    -0.80
    huk
    -0.76
     Aris
    -0.72
    scriber
    -0.69
     Bridgewater
    -0.68
    لفل
    -0.68
     sighed
    -0.67
     bundle
    -0.66
     scenario
    -0.65
     kèm
    -0.65
    POSITIVE LOGITS
     Wilder
    1.03
     wilder
    0.89
     Welles
    0.86
    Security
    0.82
     mov
    0.80
    Typical
    0.78
     قیمت
    0.76
    Directories
    0.75
    security
    0.75
     Hull
    0.74
    Activations Density 0.008%

    No Known Activations