© 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. 35313
    Prev
    Next
    INDEX
    Explanations

    implementation research paper

    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
    æī¿
    -0.10
    åĪĽä¸ļæĿ¿
    -0.09
    绵
    -0.09
    æĶ¾å®½
    -0.08
    åIJĮåѦ们
    -0.08
     circum
    -0.08
    æĬļ
    -0.08
    æ´
    -0.08
    æ³Ħ
    -0.08
     Ethan
    -0.08
    POSITIVE LOGITS
    æķĻçłĶ
    0.13
    离èģĮ
    0.11
    istrict
    0.11
    åݿ级
    0.10
    æķĻå¸Ī
    0.10
    æľĥèŃ°
    0.10
    çľģ级
    0.10
    ä¼ļè®®
    0.10
     districts
    0.10
    iders
    0.10
    Activations Density 0.136%

    No Known Activations