© 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. 3422
    Prev
    Next
    INDEX
    Explanations

    rod or rod followed by name parts

    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
    ariat
    -0.10
    enance
    -0.10
    éħĭ
    -0.10
    ãģĦãģĨ
    -0.10
    eric
    -0.09
    裳
    -0.09
    ä½Ĺ
    -0.09
    HashCode
    -0.09
    eds
    -0.08
    sic
    -0.08
    POSITIVE LOGITS
    åı¨
    0.12
    èĩ¾
    0.10
    ically
    0.10
    velt
    0.10
    yssey
    0.10
    backs
    0.09
    -shaped
    0.09
    ities
    0.09
    tails
    0.09
    cade
    0.08
    Activations Density 0.022%

    No Known Activations