© 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. 12776
    Prev
    Next
    INDEX
    Explanations

    Wild Falken

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

    The neuron fires on numeric tokens—especially standalone numbers and measured quantities (integers, decimals, 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_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
    Televis
    -1.30
     pérd
    -1.07
    Heer
    -1.05
     applau
    -1.05
     nou
    -1.04
     tendrán
    -1.03
     deberán
    -1.02
    Ramon
    -1.02
     immen
    -1.01
    蓦
    -1.01
    POSITIVE LOGITS
     also
    1.51
     still
    1.32
     some
    1.23
     not
    1.23
     what
    1.19
     those
    1.18
     Both
    1.16
     subsequently
    1.14
     your
    1.11
     just
    1.09
    Activations Density 0.053%

    No Known Activations