© 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. 33015
    Prev
    Next
    INDEX
    Explanations

    property descriptions and valuations

    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
    姥
    -0.11
    ë§IJ
    -0.11
    æ¤ħ
    -0.10
    iper
    -0.10
    éĻµ
    -0.09
    _WM
    -0.09
    ilver
    -0.09
     ÐłÐµÑģпÑĥб
    -0.09
     Thá»ĭ
    -0.09
    ingly
    -0.09
    POSITIVE LOGITS
    aceous
    0.13
    ties
    0.11
    folio
    0.10
    hecy
    0.09
    holders
    0.09
    åŃ¦å®¶
    0.09
    sons
    0.09
    (nonatomic
    0.08
    wide
    0.08
    æ¬Ĭ
    0.08
    Activations Density 0.065%

    No Known Activations