© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 79012
    Prev
    Next
    INDEX
    Explanations

    program control keywords

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

    This neuron specializes in spotting single‐word control‐flow or error-handling statements (e.g. `pass`, `quit`, `shift`) that swallow errors or advance/exit loops.

    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
    ကိုးက
    -1.05
     tanong
    -1.02
     komun
    -1.01
     bús
    -0.97
    僦
    -0.96
    Conclusión
    -0.95
    ină
    -0.95
    之间
    -0.92
    ⡰
    -0.91
     désir
    -0.90
    POSITIVE LOGITS
     skipping
    1.05
     skip
    1.02
    ;
    1.02
    h
    1.00
    m
    0.94
    ;
    0.91
    skip
    0.90
     Skip
    0.86
     delivering
    0.84
    $>$
    0.84
    Activations Density 0.002%

    No Known Activations