© 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. 72798
    Prev
    Next
    INDEX
    Explanations

    specific services and places

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

    The neuron is keyed to detecting capitalized proper names and the punctuation (commas, apostrophes) that typically surround them in running text.

    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
     aptly
    -1.57
     undeniably
    -1.43
     had
    -1.39
     set
    -1.38
     out
    -1.35
     this
    -1.35
     at
    -1.35
     ensuring
    -1.34
    ？
    -1.33
     facilitating
    -1.30
    POSITIVE LOGITS
     applau
    1.92
     célè
    1.62
     bricola
    1.48
     sirop
    1.46
    Vlast
    1.41
     télécharge
    1.41
     voient
    1.40
     citazioni
    1.38
    嬉しいです
    1.37
     reveló
    1.34
    Activations Density 0.157%

    No Known Activations