© 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. 58146
    Prev
    Next
    INDEX
    Explanations

    they say or mean

    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
    å±±åºĦ
    -0.10
    ä¸Ńæĸĩ
    -0.09
    aget
    -0.09
    缸å½ĵ
    -0.09
    ä¼´ä¾£
    -0.09
    åįĹå±±
    -0.09
    ç¢į
    -0.08
    天涯
    -0.08
    å¾ĺå¾Ĭ
    -0.08
    座
    -0.08
    POSITIVE LOGITS
    çIJ¢
    0.10
    ä¾
    0.09
    леÑĢ
    0.08
    cir
    0.08
    说æĪij
    0.08
    jang
    0.08
    soever
    0.08
    çľ¼åħī
    0.08
    åijĬè¯īä»ĸ
    0.08
    纸ä¸Ĭ
    0.08
    Activations Density 0.047%

    No Known Activations