© 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. 49881
    Prev
    Next
    INDEX
    Explanations

    seams and joints

    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
    çĶ¨åĵģ
    -0.12
    çĭIJ
    -0.09
     fram
    -0.09
    çĶ¨éĢĶ
    -0.09
    ç¿ł
    -0.09
    owan
    -0.09
    å°Ĩ以
    -0.09
    临
    -0.08
    åĥµå°¸
    -0.08
     camp
    -0.08
    POSITIVE LOGITS
    å¤Ħ
    0.14
    è¡Ķæİ¥
    0.12
    ç¼Ŀ
    0.11
    ç¼ĿéļĻ
    0.11
    åIJĪ
    0.11
     seams
    0.10
    æİ¥
    0.10
    èĻķ
    0.10
     joins
    0.09
    ä¸įåĪ°
    0.09
    Activations Density 0.102%

    No Known Activations