© 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. 48499
    Prev
    Next
    INDEX
    Explanations

    posterior or rear part

    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
    娣
    -0.11
    æĴŃ
    -0.09
    pread
    -0.09
    åīįéĢĶ
    -0.09
    otron
    -0.09
     Boyle
    -0.09
     pont
    -0.09
    èij©
    -0.08
    hest
    -0.08
     Bs
    -0.08
    POSITIVE LOGITS
     lÆ°ng
    0.14
    (back
    0.14
    .Back
    0.12
    =back
    0.12
    /back
    0.12
    尾巴
    0.11
    WARDS
    0.10
     Behind
    0.10
     atrás
    0.10
    _BACKEND
    0.10
    Activations Density 0.152%

    No Known Activations