© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 35133
    Prev
    Next
    INDEX
    Explanations

    Q: Project highlights

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

    The neuron is primarily detecting common English function words (especially the definite article “the” and similar high‐frequency stop 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
    Ᏼ
    -0.94
     our
    -0.91
     Torna
    -0.90
     Because
    -0.87
    dze
    -0.86
     discogs
    -0.82
     electrolux
    -0.82
     Psychiat
    -0.81
     kyo
    -0.81
     luxo
    -0.80
    POSITIVE LOGITS
     meanwhile
    1.02
    Meanwhile
    1.01
    }\]
    1.00
    rary
    0.99
    gesetzt
    0.96
     livrer
    0.94
    人も
    0.90
     também
    0.90
    wickshire
    0.90
     salt
    0.90
    Activations Density 0.001%

    No Known Activations