© 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. 4752
    Prev
    Next
    INDEX
    Explanations

    suffixes like -ary, -ory, -ium

    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
    ening
    -0.12
    ened
    -0.12
    æıĦ
    -0.09
    禧
    -0.09
    anim
    -0.09
    æľĢ好çļĦ
    -0.09
    оказ
    -0.09
    家åĸ»æĪ·æĻĵ
    -0.08
    æłĵ
    -0.08
    æĤ²åī§
    -0.08
    POSITIVE LOGITS
    ount
    0.09
    ello
    0.09
    tion
    0.08
    iolet
    0.08
    ECTOR
    0.08
     Dude
    0.08
     whereabouts
    0.08
    _invoke
    0.08
    oji
    0.08
    inary
    0.07
    Activations Density 0.081%

    No Known Activations