© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 75232
    Prev
    Next
    INDEX
    Explanations

    Twitter, org, Lee, 5, Q

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

    The neuron is picking out code‐like alphanumeric sequences—e.g. Base64 URL fragments and software 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_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
     their
    -1.29
     all
    -1.23
     bereits
    -1.20
     began
    -1.12
     begin
    -1.04
     get
    -1.01
    者の
    -1.01
    ParallelGroup
    -1.00
     for
    -0.99
     Área
    -0.99
    POSITIVE LOGITS
     ativar
    1.27
     connaissent
    1.25
     mesmas
    1.24
     assistir
    1.22
     actualizar
    1.20
     alugar
    1.20
     baixar
    1.18
     sinais
    1.18
     représentent
    1.16
     pensamentos
    1.16
    Activations Density 0.000%

    No Known Activations