© 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. 49705
    Prev
    Next
    INDEX
    Explanations

    author's own opinions and expectations

    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
    æĪIJ为ä¸Ģ个
    -0.10
    çķĻä¸ĭ
    -0.10
    çļĦèĥĮåIJİ
    -0.09
    .ResponseBody
    -0.09
    æĶ¾æīĭ
    -0.09
    å¤į
    -0.09
    auce
    -0.09
    å®īå¿ĥ
    -0.09
    xbe
    -0.08
    移åĪ°
    -0.08
    POSITIVE LOGITS
    ä¿ı
    0.10
    å½ĵæŶ
    0.09
    å½ĵæŶçļĦ
    0.09
    åIJĦèĩªçļĦ
    0.09
    ç¼Ģ
    0.09
    éĸ
    0.09
    ä»ħ代表
    0.08
    è¿ľæĻ¯
    0.08
    èĮ¨
    0.08
    AMES
    0.08
    Activations Density 0.012%

    No Known Activations