© 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. 33549
    Prev
    Next
    INDEX
    Explanations

    removing or reducing actions

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

    The neuron selectively responds to numerical and quantitative tokens (e.g. dosage codes, measurements) and modal/adverbial terms that indicate degree or potential (“could,” “potentially,” “immediately,” etc.).

    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
     still
    -1.23
     flexibilidad
    -1.21
     always
    -1.09
     aún
    -1.05
     hjælpe
    -1.05
     need
    -1.03
     jurídica
    -1.02
     اكتوبر
    -1.00
     should
    -1.00
     already
    -0.98
    POSITIVE LOGITS
    ลด
    1.16
     remove
    1.13
     kitu
    1.12
     rid
    1.09
     fewer
    1.05
    remove
    1.05
    agerie
    1.04
    eavor
    1.00
    Elimin
    1.00
     أقل
    0.98
    Activations Density 0.101%

    No Known Activations