© 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. 3245
    Prev
    Next
    INDEX
    Explanations

    routine application exclude universally

    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
    .Imp
    -0.11
    (pb
    -0.09
    å¹²
    -0.09
    itzer
    -0.09
     Witness
    -0.09
     Webster
    -0.09
    _keeper
    -0.09
    éļ¼
    -0.08
    .fixture
    -0.08
    é³³
    -0.08
    POSITIVE LOGITS
    ä¸Ģåij³
    0.16
    ä¸Ģå¾ĭ
    0.14
    çĽ²çĽ®
    0.14
    çĽ²
    0.13
    blind
    0.13
    ä¸ĢåĪĢ
    0.13
     blind
    0.13
    çļĦä¿¡ä»»
    0.12
     blindly
    0.12
     Blind
    0.11
    Activations Density 0.121%

    No Known Activations