© 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. 18379
    Prev
    Next
    INDEX
    Explanations

    profanity followed by specific suffixes

    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
    .usage
    -0.10
    égor
    -0.09
     Gül
    -0.09
    _FS
    -0.09
    åĬŁå¤«
    -0.09
    åİļåİļçļĦ
    -0.09
    _EOL
    -0.09
    æĭľ
    -0.09
    _EVAL
    -0.09
    ãĥ¼ãĥ³
    -0.09
    POSITIVE LOGITS
    able
    0.11
    èĢħçļĦ
    0.11
    ety
    0.10
    ening
    0.10
    ers
    0.09
    intosh
    0.09
    æĪijèĥ½
    0.08
    ably
    0.08
    èĭĦ
    0.08
    ish
    0.08
    Activations Density 0.007%

    No Known Activations