© 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. 71156
    Prev
    Next
    INDEX
    Explanations

    technical states or modifications

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

    The neuron activates on verbs that describe manipulating or processing content (e.g. add, scale, scroll, replicate, paste, blur, update).

    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
     vanish
    -0.96
     aident
    -0.93
    不但
    -0.93
     mêmes
    -0.92
    と言われて
    -0.92
     Même
    -0.88
    évalu
    -0.88
    仅仅
    -0.88
    仅
    -0.84
     vůbec
    -0.82
    POSITIVE LOGITS
     or
    1.95
     if
    1.28
     hoặc
    1.20
     или
    1.18
     atau
    1.13
     oppure
    1.11
    或者
    1.11
     then
    1.09
     could
    1.09
     during
    1.05
    Activations Density 0.066%

    No Known Activations