© 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. 9631
    Prev
    Next
    INDEX
    Explanations

    git patch commit

    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
    æĴĩ
    -0.10
     Shed
    -0.09
    å¨ħ
    -0.09
    å®ĺ
    -0.08
    ä½ľå¼Ĭ
    -0.08
    ç½ijç»ľä¼łæĴŃ
    -0.08
    èĿī
    -0.08
    èµ°
    -0.08
    éĿ´
    -0.08
    åĭ¾
    -0.08
    POSITIVE LOGITS
    ç»ıåİĨè¿ĩ
    0.09
    车èģĶç½ij
    0.09
     spur
    0.09
    ãĤ¢ãĤ¯ãĤ»
    0.09
    ä¸Ģæīĭ
    0.08
    _MAGIC
    0.08
    ç¨ĢåľŁ
    0.08
    ä¸Ģ举
    0.08
    åıijçĥ§
    0.08
    اÙĥ
    0.08
    Activations Density 0.242%

    No Known Activations