© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 73887
    Prev
    Next
    INDEX
    Explanations

    source attributions

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

    The neuron fires on all-caps source or outlet names (e.g. “LIFENEWS”)—i.e. uppercase acronyms or publication labels.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    parenthetical source or credit attributions (such as site names, links, bylines, or photo credits) in articles and posts.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    opening parentheses that introduce source attributions or citations.

    oai_token-act-pair · claude-4-5-sonnetTriggered 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
    "
    -3.05
     as
    -2.75
     with
    -2.64
     from
    -2.48
     '
    -2.41
    :
    -2.31
     "
    -2.30
     for
    -2.28
     on
    -2.25
     out
    -2.20
    POSITIVE LOGITS
     maksi
    2.42
    嵎
    2.41
    anmoins
    2.31
     fermer
    2.25
    Therapie
    2.19
     Gaulle
    2.17
     geene
    2.17
    櫺
    2.16
     kapak
    2.14
     rajut
    2.14
    Activations Density 0.025%

    No Known Activations