© 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. 58701
    Prev
    Next
    INDEX
    Explanations

    early + descriptive word

    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
    éļıæŶ
    -0.13
    éļ¨æĻĤ
    -0.10
    ä¸įåĥı
    -0.09
    å¾Īæ¸ħæ¥ļ
    -0.09
    æĥ¯
    -0.09
    UNC
    -0.09
    马ä¸Ĭ
    -0.09
    rig
    -0.09
    å½ĵä¸ĭ
    -0.09
    马ä¸Ĭå°±
    -0.09
    POSITIVE LOGITS
    -stage
    0.12
    迹象
    0.10
     childhood
    0.09
    zeitig
    0.09
    buzz
    0.09
     stages
    0.08
    å¹¼
    0.08
    赦
    0.08
    brook
    0.08
    éĢı
    0.08
    Activations Density 0.036%

    No Known Activations