© 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. 27002
    Prev
    Next
    INDEX
    Explanations

    time periods followed by noun

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

    The neuron strongly activates on descriptive adjectival modifiers (including past‐participles and adjective forms) used to qualify or describe nouns.

    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
    是怎么回事
    -1.48
     of
    -1.38
     leuke
    -1.36
     iedereen
    -1.32
    ábado
    -1.31
    戽
    -1.30
    力气
    -1.28
     grecque
    -1.27
     schützen
    -1.26
    哢
    -1.26
    POSITIVE LOGITS
     own
    2.45
     approach
    1.70
     abilities
    1.47
     attempt
    1.44
     ability
    1.44
     efforts
    1.36
     experience
    1.33
     which
    1.32
     choice
    1.24
     array
    1.21
    Activations Density 0.041%

    No Known Activations