© 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. 18018
    Prev
    Next
    INDEX
    Explanations

    punctuation followed by a word

    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
    otti
    -0.09
    äººåľ¨
    -0.08
     collo
    -0.08
    åľ¨è¿ĩ渡
    -0.08
    éĻĢ
    -0.08
    er
    -0.08
    éĤ£æĺ¯
    -0.08
     Railway
    -0.08
    çŃ¾
    -0.08
     ÑĩаÑģÑĤноÑģÑĤи
    -0.07
    POSITIVE LOGITS
    uding
    0.09
    verages
    0.09
    ngine
    0.09
    åĽŀäºĭ
    0.09
    ãģĵãģĵãģ¾ãģ§
    0.09
    ecure
    0.08
    究竣
    0.08
    ditor
    0.08
    çĿ«
    0.08
    å¦ĤæѤ
    0.08
    Activations Density 0.330%

    No Known Activations