© 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. 34-GEMMASCOPE-RES-131K
    7. 30379
    Prev
    Next
    INDEX
    Explanations

    Nou, Ant, Pol, à

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

    This neuron fires on the short, all-letter fragments used as localization keys or abbreviations (e.g. language codes like “ca,” month names like “Jan,” “Febru,” etc., and UI labels).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     pavo
    -1.19
     úl
    -1.15
     portugu
    -1.14
    たって
    -1.13
     oport
    -1.13
     emblema
    -1.13
     púb
    -1.11
    歿
    -1.10
     wouldnt
    -1.09
     dolorosa
    -1.09
    POSITIVE LOGITS
    に入れて
    1.23
    に入れる
    1.18
    に入れ
    1.08
     riusc
    1.07
    様な
    1.06
     parceria
    1.06
     laranja
    1.05
    ところは
    1.04
     costitu
    1.03
    ところです
    1.02
    Activations Density 0.001%

    No Known Activations