© 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. 38664
    Prev
    Next
    INDEX
    Explanations

    scroll down

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

    The neuron strongly fires on UI navigation or content‐movement instructions—especially “scroll,” “navigate,” “down,” and similar action 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
     when
    -1.18
     aastal
    -1.09
    ↵↵↵↵↵↵↵
    -1.02
     ALWAYS
    -1.01
     NOTHING
    -0.99
    ↵↵↵↵↵↵↵↵↵↵↵↵↵
    -0.99
    ↵↵↵↵↵
    -0.98
     edn
    -0.97
    ystis
    -0.96
     EVERYONE
    -0.96
    POSITIVE LOGITS
     into
    1.84
     the
    1.59
     to
    1.48
     page
    1.23
     jusqu
    1.20
     past
    1.18
    eſt
    1.17
    orgeous
    1.16
     ۱۶
    1.16
     весен
    1.13
    Activations Density 0.029%

    No Known Activations