© 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. 34890
    Prev
    Next
    INDEX
    Explanations

    salt and dissolve

    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
    æĸ¯åŁº
    -0.10
    å·¥ç¨ĭæľºæ¢°
    -0.09
    _CI
    -0.09
    noxious
    -0.09
    çķª
    -0.09
    ottle
    -0.09
    /rss
    -0.09
    æĸ°ä¸ī
    -0.09
    çŀª
    -0.09
    翳
    -0.09
    POSITIVE LOGITS
     salt
    0.13
    salt
    0.12
    çĽIJ
    0.12
    6
    0.11
    7
    0.11
    5
    0.11
    3
    0.10
    8
    0.10
    1
    0.10
     Salt
    0.10
    Activations Density 0.050%

    No Known Activations