© 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. 63489
    Prev
    Next
    INDEX
    Explanations

    ph- followed by common prefixes

    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
    Ú¾
    -0.10
    份é¢Ŀ
    -0.10
    INI
    -0.09
    à¹īว
    -0.09
    idad
    -0.09
     sÃłng
    -0.09
    flare
    -0.09
    igin
    -0.09
    .band
    -0.09
    zte
    -0.09
    POSITIVE LOGITS
    oenix
    0.13
    ãģĹãģ¾ãģĨ
    0.11
    ysical
    0.10
    rases
    0.10
    teenth
    0.09
    onium
    0.09
    araoh
    0.09
    åĬĽæ±Ĥ
    0.09
    ysics
    0.09
    à¹ģà¸Ĺ
    0.09
    Activations Density 0.175%

    No Known Activations