© 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. 32812
    Prev
    Next
    INDEX
    Explanations

    as expected, intended, designed

    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
    æĭĺ
    -0.09
    	Runtime
    -0.09
    çĤ·
    -0.08
    (Runtime
    -0.08
    åħ³åı£
    -0.08
    以ä¸ĬçļĦ
    -0.08
    çįŃ
    -0.08
    åĨ²
    -0.08
     اÙĦتاب
    -0.08
    ä¸įæĥľ
    -0.08
    POSITIVE LOGITS
    é¢ĦæľŁ
    0.24
     expected
    0.19
    expected
    0.17
    é¢ĦæĸĻ
    0.15
    é¢Ħ计
    0.15
    Expected
    0.14
    .expected
    0.14
    EXPECTED
    0.14
    æľŁå¾ħ
    0.14
    æľŁæľĽ
    0.13
    Activations Density 0.112%

    No Known Activations