© 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. 16885
    Prev
    Next
    INDEX
    Explanations

    say or goodbye phrases

    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
    çĭ¡
    -0.09
    çĽ¾
    -0.09
    راÙĤ
    -0.09
    uxe
    -0.09
    éªĩ
    -0.09
    ä¸Ī夫
    -0.08
    ç»ĥ
    -0.08
    æĦŁ
    -0.08
    .bb
    -0.08
    ÑĢаÐ
    -0.08
    POSITIVE LOGITS
    åijĬåĪ«
    0.16
    æijĴ
    0.13
     goodbye
    0.13
    æĭĴç»Ŀ
    0.11
    åĨįè§ģ
    0.11
     Bye
    0.10
    bye
    0.10
     farewell
    0.10
     bye
    0.09
     Say
    0.09
    Activations Density 0.025%

    No Known Activations