© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 101428
    Prev
    Next
    INDEX
    Explanations

    new books, koch county, networks

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

    The neuron fires on uppercase labels and identifiers in code‐comment headers—e.g. section titles like “TEMPLATE,” “CWE,” “Bad Source,” “Stack_Based_Buffer_Overflow,” etc.

    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
    möglichen
    -1.51
    atize
    -1.42
    utilized
    -1.41
    絝
    -1.37
    稞
    -1.37
    odyear
    -1.36
    乄
    -1.36
    傒
    -1.35
    ulate
    -1.34
    
    -1.34
    POSITIVE LOGITS
    These
    1.58
     will
    1.40
    ’
    1.38
    これらの
    1.35
     these
    1.34
     on
    1.33
     uses
    1.32
    …
    1.30
    0
    1.30
    in
    1.27
    Activations Density 0.010%

    No Known Activations