© 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. 47668
    Prev
    Next
    INDEX
    Explanations

    distributed under license

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

    The neuron strongly activates on numeric tokens—especially multi‐digit sequences like years, proposition numbers, and version numbers.

    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
     VIN
    -0.68
    bia
    -0.67
    Xi
    -0.67
    étation
    -0.66
     Saved
    -0.66
     IOS
    -0.66
     Signatures
    -0.65
     AUD
    -0.65
     einstellen
    -0.65
     ميد
    -0.65
    POSITIVE LOGITS
    oluble
    0.69
    bellar
    0.66
     milli
    0.66
    折
    0.65
    𝑍
    0.65
     naturales
    0.65
    paddingRight
    0.64
    顺利
    0.64
     totale
    0.63
     przepis
    0.62
    Activations Density 0.089%

    No Known Activations