© 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. 22-GEMMASCOPE-RES-131K
    7. 68361
    Prev
    Next
    INDEX
    Explanations

    proper nouns and code packages

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

    The neuron fires on code‐identifier tokens—especially segments of Java package and class names (e.g. “com”, “yixiang”, “QueryParam”) and related punctuation in import/package lines.

    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
    \#:
    -1.30
    🔗
    -1.23
    ❄
    -1.23
    クリル
    -1.16
    *",
    -1.15
    distingu
    -1.14
     carburant
    -1.14
     Mě
    -1.13
    ctober
    -1.12
    ÖN
    -1.12
    POSITIVE LOGITS
     of
    1.18
     up
    1.02
     좋
    0.95
     with
    0.93
    ransition
    0.90
     new
    0.90
    如果你
    0.89
     on
    0.89
    サキ
    0.88
     называют
    0.88
    Activations Density 0.001%

    No Known Activations