© 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. 26408
    Prev
    Next
    INDEX
    Explanations

    numbers and punctuation

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

    The neuron fires on structural or formatting tokens—especially punctuation marks (like commas), numbers, and other non‐word separators—rather than actual words.

    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
     its
    -1.16
    daily
    -1.11
     and
    -1.05
     Their
    -1.05
    ఢ
    -1.04
     blisko
    -0.97
     I
    -0.96
     Daily
    -0.95
     New
    -0.95
     Bewer
    -0.93
    POSITIVE LOGITS
    УДК
    1.23
     boya
    1.22
     aspi
    1.20
     pulsa
    1.20
     pasto
    1.18
     poliuret
    1.16
     identifica
    1.16
    Lundi
    1.14
     paleta
    1.14
     aspira
    1.14
    Activations Density 0.006%

    No Known Activations