© 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. 54186
    Prev
    Next
    INDEX
    Explanations

    worse

    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
    à¹Į
    -0.12
    ssa
    -0.12
    ized
    -0.10
    shire
    -0.10
    sWith
    -0.10
    ìį¨
    -0.10
    ships
    -0.10
    uxe
    -0.10
    -esque
    -0.09
    sson
    -0.09
    POSITIVE LOGITS
     gì
    0.10
    äºĨå¾Īå¤ļ
    0.10
    ened
    0.10
    -than
    0.10
    idge
    0.10
    PERT
    0.09
    ä¹İ
    0.09
    ening
    0.09
    åѦåİĨ
    0.09
    ä¸ŃéĢĶ
    0.09
    Activations Density 0.006%

    No Known Activations