© 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-3-4B-IT
    5. 29-GEMMASCOPE-2-RES-16K
    7. 3319
    Prev
    Next
    INDEX
    Explanations

    will attempt

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-4b-it/resid_post/layer_29_width_16k_l0_medium
    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
     سين
    0.57
    টিল
    0.56
    壺
    0.55
     পড়তে
    0.55
    burner
    0.55
    훤
    0.54
    縝
    0.54
    ధీ
    0.54
    ത്തില
    0.53
     asuntos
    0.53
    POSITIVE LOGITS
     attempt
    5.35
     try
    5.30
     attempts
    5.24
     tries
    5.21
     trying
    5.05
    try
    4.99
     Attempt
    4.95
    Try
    4.90
     Try
    4.90
     attempting
    4.81
    Activations Density 1.991%

    No Known Activations