© 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. 93215
    Prev
    Next
    INDEX
    Explanations

    Lozouet, Heizenrader, Rushchanski, Wetzel, Qwabe

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

    This neuron strongly activates on capitalized personal names (especially surnames) in author, credit, or taxonomy attributions.

    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
    젖
    -0.99
    حدى
    -0.92
    どうぞ
    -0.90
     similar
    -0.89
     ultimately
    -0.89
     worthwhile
    -0.87
     great
    -0.85
    zeption
    -0.84
     setCode
    -0.81
     easily
    -0.79
    POSITIVE LOGITS
    Funny
    0.98
    完善
    0.91
     Funny
    0.89
    plain
    0.88
    Conditional
    0.85
     CCP
    0.84
    钛
    0.84
    conditional
    0.84
    LOAT
    0.84
     huo
    0.83
    Activations Density 0.107%

    No Known Activations