© 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. 3246
    Prev
    Next
    INDEX
    Explanations

    wh followed by "ilst", "acky", "istles", "ine", "iten", "orning"

    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
    Çļ
    -0.10
    гла
    -0.09
    èªĵè¨Ģ
    -0.09
    ORS
    -0.09
     Symbols
    -0.08
    bursement
    -0.08
    istas
    -0.08
    andel
    -0.08
    oral
    -0.08
     shar
    -0.08
    POSITIVE LOGITS
     wh
    0.14
    -wh
    0.14
    wh
    0.13
     Wh
    0.13
    _WH
    0.13
    .wh
    0.12
    Wh
    0.11
     WH
    0.10
    _wh
    0.10
    éĴĹ
    0.09
    Activations Density 0.007%

    No Known Activations