© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26
    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 127863
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on strong, content-bearing verbs (especially past participles or gerunds) that describe actions or features—e.g. “showcased,” “rendered,” “stowing,” “access,” “bolstering”—often used in promotional or product descriptions.

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408

    past events or outcomes

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    Negative Logits
     after
    -1.49
     using
    -1.43
     or
    -1.28
     without
    -1.27
     so
    -1.23
     but
    -1.20
     unusual
    -1.20
     when
    -1.19
     one
    -1.19
     on
    -1.19
    POSITIVE LOGITS
    aarrggbb
    1.41
    oma
    1.34
    },
    1.34
     ผู้
    1.33
     iNdEx
    1.31
    ai
    1.30
    ota
    1.30
    粜
    1.25
    s
    1.25
    hlten
    1.23
    Activations Density 0.050%

    No Known Activations