© 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. 22-GEMMASCOPE-RES-131K
    7. 26684
    Prev
    Next
    INDEX
    Explanations

    field names or object types

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

    The neuron activates on uppercase‐starting identifier tokens that denote types or classes (e.g. Stream, Writer, Json, Configuration, ReadWriteLock, CallbackManager).

    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
    畀
    -1.02
    —
    -1.02
     می‌گوید
    -0.99
    :
    -0.92
     uphe
    -0.91
    lands
    -0.90
    齣
    -0.90
     этих
    -0.90
    じゃ
    -0.90
    faire
    -0.89
    POSITIVE LOGITS
     gamla
    1.44
     dipakai
    1.20
    ᅠ
    1.16
    thaf
    1.13
     clube
    1.12
    pantal
    1.11
     flesta
    1.10
    ceptable
    1.08
     pacient
    1.08
    lorn
    1.07
    Activations Density 0.017%

    No Known Activations