© 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. 49821
    Prev
    Next
    INDEX
    Explanations

    weight, size, or shape descriptions

    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
    itive
    -0.09
    å±ĤéĿ¢
    -0.09
    abouts
    -0.09
    sWith
    -0.08
    ald
    -0.08
    ubble
    -0.08
    rior
    -0.08
    Off
    -0.08
    丰å¯Įå¤ļ彩
    -0.08
    rup
    -0.08
    POSITIVE LOGITS
    /he
    0.10
    ming
    0.10
    ifying
    0.10
    fony
    0.10
     Chance
    0.10
    est
    0.10
    人èĦī
    0.10
    atures
    0.10
    ãģ£ãģ¨
    0.10
    advert
    0.09
    Activations Density 0.007%

    No Known Activations