© 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. 48894
    Prev
    Next
    INDEX
    Explanations

    decide, looked, mentioned, and then

    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
    \↵
    -0.12
    ãģĿãĤĮãģŀ
    -0.11
    ä¸İæĹ¶ä¿±
    -0.10
    ä¹łè¿ijå¹³æĢ»
    -0.10
    ãĤ¨ãĥįãĥ«
    -0.09
    "\↵
    -0.09
    ãģ©ãĤĵãģ©
    -0.09
     misunder
    -0.08
    æĥ¯
    -0.08
    èĦ±é¢ĸ
    -0.08
    POSITIVE LOGITS
    å¹´å¤ľ
    0.11
    NewLabel
    0.09
    aura
    0.09
    iscard
    0.09
    leanup
    0.09
    %E
    0.08
    Âł
    0.08
    erp
    0.08
    ystals
    0.08
    theValue
    0.08
    Activations Density 0.265%

    No Known Activations