© 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-3-27B-IT
    5. 41-GEMMASCOPE-2-RES-262K
    7. 5076
    Prev
    Next
    INDEX
    Explanations

    collisions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post_all/layer_41_width_262k_l0_big
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ามารถ
    0.47
     utilisent
    0.46
    conjugated
    0.46
    ratingBarStyle
    0.45
     রাখে
    0.44
     nhu
    0.44
    癢
    0.44
     arcu
    0.43
    িক্ত
    0.43
    移行
    0.42
    POSITIVE LOGITS
     collision
    2.14
    Collision
    1.91
     collisions
    1.91
     collide
    1.89
     Collision
    1.89
    collision
    1.86
    碰撞
    1.80
     crash
    1.78
     collided
    1.78
    撞
    1.75
    Activations Density 0.007%

    No Known Activations