© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 11502
    Prev
    Next
    INDEX
    Explanations

    capture and being taken

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    asure
    -0.10
    涨
    -0.09
    otton
    -0.08
    帷
    -0.08
    akening
    -0.08
    è¾¼ãģ¿
    -0.08
    èĿĻ
    -0.08
    coat
    -0.08
     murderers
    -0.08
    itten
    -0.08
    POSITIVE LOGITS
    æݳ
    0.13
    ä¿ĺ
    0.12
     capture
    0.11
    èĻı
    0.11
     captured
    0.11
    .capture
    0.10
    Capture
    0.10
     Capture
    0.10
     captures
    0.10
    æİł
    0.10
    Activations Density 0.036%

    No Known Activations