© 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. 47456
    Prev
    Next
    INDEX
    Explanations

    No

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

    The neuron fires on tokens that come from programming‐API or code contexts (e.g. API/class names, code symbols, numeric literals, version or release identifiers)—in other words, it spots code and technical-API terminology.

    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
    license
    -0.73
    lais
    -0.70
    frm
    -0.69
     đua
    -0.69
     licenses
    -0.67
    ummings
    -0.65
    ্য
    -0.64
     مع
    -0.64
     Guillaume
    -0.64
    ulsory
    -0.64
    POSITIVE LOGITS
     娜
    0.66
    པའི་
    0.65
    panda
    0.65
     Quasi
    0.64
    BOSS
    0.64
    =[[
    0.63
    žová
    0.63
    𝐚
    0.62
     Naive
    0.62
     boucles
    0.62
    Activations Density 0.095%

    No Known Activations