© 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. 102734
    Prev
    Next
    INDEX
    Explanations

    principal findings and results

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

    This neuron responds to structural document markers—section headings (e.g. “Methodology,” “Principal Findings”), punctuation like “/,” and numbers—i.e. tokens used in outlines, headings, or numeric data.

    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
    redge
    -0.80
    ubs
    -0.79
     dueño
    -0.77
    фом
    -0.76
    ofu
    -0.76
     hoort
    -0.75
    Nel
    -0.73
    att
    -0.73
     Dez
    -0.72
     Quin
    -0.71
    POSITIVE LOGITS
     findings
    7.47
     Findings
    5.34
    Findings
    5.31
     finding
    4.44
     FINDINGS
    4.34
    find
    3.72
    FIND
    3.41
     finds
    3.27
    finding
    3.27
     hallaz
    3.23
    Activations Density 0.150%

    No Known Activations