© 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. 29936
    Prev
    Next
    INDEX
    Explanations

    can’t undo or fix

    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
    åı£æ°Ķ
    -0.09
    çĤ¯
    -0.09
     Fel
    -0.08
    éĥ½è¯´
    -0.08
    çľĭæĿ¥
    -0.08
    å¿«æĬ¥
    -0.08
    åIJİå¤ĩ
    -0.08
    éľ²
    -0.08
    acity
    -0.08
    è¿·
    -0.08
    POSITIVE LOGITS
    åĪĽéĢłåĩº
    0.10
    _SRC
    0.09
    anken
    0.09
    妨
    0.09
    æĵĢ
    0.09
    rets
    0.09
    伤çĹħ
    0.09
    .instant
    0.09
    éªij
    0.09
     pretending
    0.08
    Activations Density 0.055%

    No Known Activations