© 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. 31307
    Prev
    Next
    INDEX
    Explanations

    geometry and shapes

    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
    æĭį
    -0.08
     perk
    -0.08
    念
    -0.08
    kins
    -0.08
    çŃĴ
    -0.08
    aul
    -0.07
    7
    -0.07
    0
    -0.07
     SEL
    -0.07
    3
    -0.07
    POSITIVE LOGITS
    éĴĿ
    0.12
    .Apis
    0.11
    ä¸Ģè¾¹
    0.11
    è¾¹
    0.11
     cạnh
    0.10
     ÑģÑĤоÑĢонÑĭ
    0.10
    éĢĤæŶ
    0.10
     lengths
    0.10
    第ä¸ĢæĿ¡
    0.10
    éĶIJ
    0.10
    Activations Density 0.077%

    No Known Activations