© 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. 2457
    Prev
    Next
    INDEX
    Explanations

    instead, rather, or but instead

    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
    éĴ¹
    -0.10
    à¹ģà¸ģ
    -0.10
    仲
    -0.10
    å¯Ŀ
    -0.09
    æĬ¤
    -0.09
    å¹²äºĭ
    -0.09
    没人
    -0.09
    å¥ĸéĩij
    -0.09
    å¹´æĿ¥
    -0.09
    ä¸įæŃ¢
    -0.09
    POSITIVE LOGITS
    essa
    0.10
    ä¸Ģå¾ĭ
    0.10
    åĢį
    0.09
    nor
    0.08
    èĢĮæĺ¯
    0.08
    å§ĵ
    0.08
    é¹Ń
    0.08
     Lowe
    0.08
    骧
    0.08
    èĭĹ
    0.08
    Activations Density 0.037%

    No Known Activations