© 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. 34433
    Prev
    Next
    INDEX
    Explanations

    something doesn't work

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

    The neuron consistently activates only on the single character “t” token wherever it appears.

    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
    ptonshire
    -1.30
     both
    -1.29
     dakika
    -1.27
     stik
    -1.27
    參考文獻
    -1.21
    اژ
    -1.20
     temi
    -1.20
     zaidi
    -1.20
     cdti
    -1.17
     ApiService
    -1.16
    POSITIVE LOGITS
    感じが
    1.63
     necessarily
    1.63
    って何
    1.59
    に行ってきました
    1.57
    になりたい
    1.46
    為に
    1.46
    気がします
    1.44
    様な
    1.42
    ができます
    1.41
    感じです
    1.41
    Activations Density 0.045%

    No Known Activations