© 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. 49005
    Prev
    Next
    INDEX
    Explanations

    splits off or veer off

    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
    麻
    -0.10
    ergy
    -0.09
     nip
    -0.09
    pose
    -0.08
    éĿ
    -0.08
    æĸ°éĻĪ代谢
    -0.08
    ãĥĭ
    -0.08
    inel
    -0.08
    _hd
    -0.08
    èĵī
    -0.08
    POSITIVE LOGITS
     himself
    0.12
    ä»ĸèĩªå·±
    0.11
    ä»ĸ认为
    0.11
    åľ¨ä»ĸçļĦ
    0.10
    ä»ĸçļĦ
    0.10
    ä»ĸåľ¨
    0.10
    åĴĮä»ĸçļĦ
    0.09
    å®ĥçļĦ
    0.09
    çĿĢä»ĸ
    0.09
    绿水
    0.09
    Activations Density 0.263%

    No Known Activations