© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 79933
    Prev
    Next
    INDEX
    Explanations

    numbers in parentheses

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

    The neuron strongly activates on technical labels and codes—uppercase acronyms, numbered identifiers, and similar tokens (e.g. PDA, NXX, C2, station_coord, projections).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     by
    -1.21
     from
    -1.04
     with
    -1.01
     on
    -0.98
     in
    -0.91
     around
    -0.89
    付属品
    -0.89
     other
    -0.88
     at
    -0.86
     Other
    -0.85
    POSITIVE LOGITS
    ’.
    1.06
    }.
    1.05
    这样
    1.00
    geschoss
    1.00
    switch
    1.00
    kana
    0.93
    这种
    0.93
    LikeLike
    0.92
    baden
    0.92
     ainsi
    0.90
    Activations Density 0.102%

    No Known Activations