© 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. 33570
    Prev
    Next
    INDEX
    Explanations

    negating actions or descriptions

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

    The neuron spots negation and exclusion words—especially “not” and similar negative/disclaimer constructions.

    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
    分别是
    -1.08
    Bises
    -1.03
    нятие
    -0.99
    Appropriate
    -0.98
    нительно
    -0.97
     atleast
    -0.96
    jnych
    -0.95
    both
    -0.93
     riesce
    -0.93
    appropriate
    -0.91
    POSITIVE LOGITS
     necessarily
    1.38
     only
    1.19
     только
    1.09
    只是
    1.07
     actual
    1.05
     but
    1.05
     merely
    1.00
    “
    0.97
     新作
    0.96
    直接
    0.95
    Activations Density 0.057%

    No Known Activations