© 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. 22496
    Prev
    Next
    INDEX
    Explanations

    casting shadows, shadow-truths

    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
     autos
    -0.09
    æŃĮ
    -0.09
    ç³ľ
    -0.08
     reb
    -0.08
     Rage
    -0.08
     Gast
    -0.08
    潺
    -0.08
    Gas
    -0.08
    æĹłéĻIJ
    -0.08
    èĢ³
    -0.08
    POSITIVE LOGITS
    éĺ´å½±
    0.26
     shadow
    0.25
     shadows
    0.24
    å½±
    0.24
    shadow
    0.22
     Shadows
    0.19
    Shadow
    0.19
    _shadow
    0.19
    .shadow
    0.19
     Shadow
    0.18
    Activations Density 0.044%

    No Known Activations