© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 29676
    Prev
    Next
    INDEX
    Explanations

    forced removal or departure

    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
    Negative Logits
    lessly
    -0.11
    ness
    -0.11
    åįģ足
    -0.10
    elves
    -0.10
    çĿ¨
    -0.10
    ãģĨãģ¡ãģ«
    -0.09
    ograf
    -0.09
    ings
    -0.09
    ä¼ļè§īå¾Ĺ
    -0.09
    lessness
    -0.09
    POSITIVE LOGITS
    ively
    0.10
    ible
    0.09
    ãĥ¼ãĤ·ãĥ§ãĥ³
    0.09
    apo
    0.09
    ees
    0.09
    apel
    0.09
    äºĨä¸Ģæī¹
    0.09
    ä¸Ģæī¹
    0.09
    eer
    0.09
    اÙĥ
    0.09
    Activations Density 0.051%

    No Known Activations