© 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. 33188
    Prev
    Next
    INDEX
    Explanations

    questions, statements, or parts of "I've"

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

    This neuron detects the introductory phrases of individual riddle or puzzle clues (e.g. lines like “Something a horse says,” “The largest living land animal,” 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_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
     escaleras
    -1.16
    iation
    -1.12
     these
    -1.10
     cordón
    -1.05
     coba
    -1.03
     Gegend
    -1.02
     sogenannten
    -0.98
    
    -0.98
     aforementioned
    -0.97
    ↔
    -0.95
    POSITIVE LOGITS
     ďal
    1.10
    𝐘
    1.09
     マンション
    1.07
    美味しく
    0.99
    agissait
    0.98
    tható
    0.97
     鸟
    0.97
     alguien
    0.97
     миллионов
    0.96
    客様
    0.95
    Activations Density 0.001%

    No Known Activations