© 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. 58571
    Prev
    Next
    INDEX
    Explanations

    manipulation or manipulate

    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
    owing
    -0.12
    aller
    -0.10
    ë¹Ļ
    -0.10
     dÃłng
    -0.09
    aday
    -0.09
    رÙĪÙħ
    -0.09
    è¡·
    -0.09
    chant
    -0.09
    ãģĦãĤĭ
    -0.09
    lix
    -0.09
    POSITIVE LOGITS
    èĩªå¦Ĥ
    0.10
     manip
    0.10
    är
    0.09
    orth
    0.09
    /man
    0.09
    uten
    0.09
    ively
    0.09
    ç¿Ĭ
    0.09
    plier
    0.09
     Toy
    0.09
    Activations Density 0.023%

    No Known Activations