© 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. 10-GEMMASCOPE-RES-131K
    7. 14819
    Prev
    Next
    INDEX
    Explanations

    ./ followed by file or command

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

    The neuron responds to “dot-” punctuation sequences in code—especially “./” path prefixes and “.$” string-concatenation operators.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     "
    -2.50
     
    -2.50
     были
    -2.44
    begin
    -2.38
    Such
    -2.30
    THE
    -2.27
    self
    -2.27
    Good
    -2.13
    section
    -2.06
    さん
    -2.05
    POSITIVE LOGITS
    羕
    3.19
    祧
    2.72
     gleichen
    2.67
    釭
    2.59
    dámské
    2.52
     Leuten
    2.48
     chociaż
    2.48
     mauva
    2.48
     selben
    2.44
    蚨
    2.39
    Activations Density 0.004%

    No Known Activations