© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 124576
    Prev
    Next
    INDEX
    Explanations

    not straightforwardly but

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

    The neuron fires on emphatic or generalizing qualifiers—words and short phrases that assert broad, absolute, or statistical conditions (e.g. “not entirely random,” “one size fits all,” “remain stationary”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     not
    -1.17
     and
    -1.15
    (_)
    -1.11
    leau
    -1.11
    าศ
    -1.06
     if
    -1.05
    𝘁
    -1.05
    ച്
    -1.02
    sss
    -1.01
     properly
    -0.99
    POSITIVE LOGITS
     receita
    1.19
    もありました
    1.16
    dieron
    1.13
    encrypted
    1.13
     épices
    1.10
     some
    1.08
     militares
    1.08
    Encrypted
    1.05
    Garrett
    1.05
     polé
    1.05
    Activations Density 0.071%

    No Known Activations