© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 7296
    Prev
    Next
    INDEX
    Explanations

    than as

    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
    åĪ«çļĦ
    -0.11
    KeyName
    -0.10
    ä¸ĢéĹª
    -0.10
    udent
    -0.09
    lica
    -0.09
    orno
    -0.09
    nero
    -0.08
     ted
    -0.08
    kinson
    -0.08
    ila
    -0.08
    POSITIVE LOGITS
    èİ«è¿ĩäºİ
    0.13
    äºİæѤ
    0.11
    adic
    0.11
    æĿ¥è¿ĻéĩĮ
    0.10
    è¿ĻéĩĮçļĦ
    0.10
    è¿Ļ家
    0.09
    eme
    0.09
    ãģĤãĤĮ
    0.08
    igy
    0.08
    ä¸ĬäºĨ
    0.08
    Activations Density 0.015%

    No Known Activations