© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 3596
    Prev
    Next
    INDEX
    Explanations

    Air crashed

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

    The neuron fires strongly on the first “substantive” token in each new document segment—i.e. it flags the opening content word or header token that begins a block.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     loadData
    -0.74
     via
    -0.71
     totally
    -0.71
     Odisha
    -0.70
    Elemental
    -0.70
     anonymity
    -0.70
     Germany
    -0.70
     jsonData
    -0.69
     motivasi
    -0.69
     using
    -0.68
    POSITIVE LOGITS
    Commun
    0.97
    τρ
    0.90
    WriteLiteral
    0.88
    क्या
    0.85
    onias
    0.84
    encoj
    0.84
    Margin
    0.84
    ‌.
    0.83
     devrait
    0.83
    quea
    0.82
    Activations Density 0.000%

    No Known Activations