© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 32010
    Prev
    Next
    INDEX
    Explanations

    jump

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

    The neuron is keyed to “header”/metadata-style tokens and markup directives (e.g. HTTP header names like “Accept-Language” or “Content-type,” comment-or-preprocessor markers like “#,” etc.).

    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
     those
    -1.00
     started
    -0.96
    glam
    -0.96
    pneu
    -0.90
     within
    -0.88
     these
    -0.86
     with
    -0.85
     dimas
    -0.85
     studies
    -0.84
    accessories
    -0.84
    POSITIVE LOGITS
    jump
    1.80
    Jump
    1.79
     jump
    1.68
     Jump
    1.63
    JUMP
    1.45
     jumping
    1.38
    ジャンプ
    1.35
     JUMP
    1.23
     jumps
    1.22
    跳
    1.21
    Activations Density 0.000%

    No Known Activations