© 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. 23145
    Prev
    Next
    INDEX
    Explanations

    erasing, Ernst, ermine, Ernt, Erase, measurement, Errol

    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
    sWith
    -0.10
    ität
    -0.09
    åĪĩãĤĬ
    -0.09
    verständlich
    -0.09
    ÙĪÙĩ
    -0.09
    iliary
    -0.09
    eur
    -0.08
    iliar
    -0.08
    autor
    -0.08
    ourn
    -0.08
    POSITIVE LOGITS
    aceous
    0.10
    neck
    0.10
    statt
    0.10
    ormsg
    0.09
    iginal
    0.09
    æİīäºĨ
    0.09
    zeug
    0.09
    ding
    0.09
    æİī
    0.09
    riott
    0.09
    Activations Density 0.038%

    No Known Activations