© 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. 22-GEMMASCOPE-RES-131K
    7. 72168
    Prev
    Next
    INDEX
    Explanations

    (a), 2), inset, Source, Below, nation, Stim

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

    This neuron fires on figure or panel labels and related caption cues (e.g. “(a)”, “A”, “Top”, “Below”, “left”, “right”, “inset”), i.e. tokens used to reference parts of a figure.

    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
     ។
    -1.25
     möglicherweise
    -1.14
     будут
    -1.09
     permitirá
    -1.07
     μπορεί
    -1.07
     berühm
    -1.03
     akár
    -1.03
    ritsar
    -1.03
     mencionado
    -1.01
    獗
    -1.01
    POSITIVE LOGITS
    ,
    1.60
     and
    1.27
     is
    1.17
    )
    1.13
     murale
    1.05
    <b>
    1.00
     latine
    0.99
     shows
    0.97
    Vapour
    0.97
     zoll
    0.97
    Activations Density 0.002%

    No Known Activations