© 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. 61490
    Prev
    Next
    INDEX
    Explanations

    copyright year and author

    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
    è¿ĽæŃ¥
    -0.10
    æįĮ
    -0.10
    pile
    -0.09
    веÑĢÑħ
    -0.09
    饴
    -0.09
    eric
    -0.08
    ä¸Ģå®ļçļĦ
    -0.08
    łĢ
    -0.08
    t
    -0.08
    eros
    -0.08
    POSITIVE LOGITS
    eenth
    0.24
     migrationBuilder
    0.15
     addCriterion
    0.13
    een
    0.12
    деÑģÑıÑĤ
    0.12
    ми
    0.11
    ozilla
    0.10
    ä¹Ŀåįģ
    0.10
    ths
    0.10
    th
    0.10
    Activations Density 0.565%

    No Known Activations