© 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. 38363
    Prev
    Next
    INDEX
    Explanations

    below code or formula

    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
    以ä¸ĬçļĦ
    -0.10
    åīįéĢĶ
    -0.10
    ä¸Ĭè¿°
    -0.10
    fully
    -0.10
     above
    -0.09
    èµ·æĿ¥
    -0.09
    naire
    -0.09
    è¿Ļéĥ¨åĪĨ
    -0.09
    ä¸Ĭä¸ĭ游
    -0.09
    以ä¸Ĭ
    -0.09
    POSITIVE LOGITS
     dozen
    0.10
    -described
    0.09
    inate
    0.09
    Ñĩик
    0.08
    ä¸įè¿ľå¤Ħ
    0.08
    æĺ¯æĪijçļĦ
    0.08
    -sex
    0.08
    éd
    0.08
    chant
    0.08
    oque
    0.08
    Activations Density 0.041%

    No Known Activations