© 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. 34-GEMMASCOPE-RES-131K
    7. 25738
    Prev
    Next
    INDEX
    Explanations

    commercial license

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

    The neuron responds most strongly to uncommon or content-specific tokens (i.e. low-frequency 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_34/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
     so
    -1.13
     although
    -1.10
     Undoubtedly
    -1.00
     Discard
    -0.99
     bane
    -0.95
    人間の
    -0.93
     ефек
    -0.92
    しております
    -0.92
    Edd
    -0.91
    に到着
    -0.91
    POSITIVE LOGITS
     ♪
    1.30
     kuin
    1.24
    ♪
    1.23
     kinda
    1.18
     kiti
    1.17
     financi
    1.12
    confirm
    1.09
    vention
    1.09
    provide
    1.09
    new
    1.05
    Activations Density 0.003%

    No Known Activations