© 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. 56831
    Prev
    Next
    INDEX
    Explanations

    off-topic or duplicate discussions

    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
    entai
    -0.12
    оÑı
    -0.10
    imson
    -0.10
    å¾ħ
    -0.09
    lightbox
    -0.09
    ôm
    -0.09
     Sext
    -0.09
    æ±°
    -0.09
    æłĹ
    -0.09
    å¾Ļ
    -0.09
    POSITIVE LOGITS
    pus
    0.09
    3
    0.09
    eness
    0.08
    4
    0.08
    1
    0.08
    8
    0.08
    6
    0.08
    chod
    0.08
    åħį
    0.08
    9
    0.07
    Activations Density 0.004%

    No Known Activations