© 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. 55043
    Prev
    Next
    INDEX
    Explanations

    ghosts and spirits

    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
    æĥħ趣
    -0.10
    æľ¬ç½ijç«Ļ
    -0.10
    积æŀģæİ¨è¿Ľ
    -0.10
    鹬
    -0.10
    alcon
    -0.09
    麻çĹ¹
    -0.09
    庸
    -0.09
    éĺIJ
    -0.09
    çĶ¥
    -0.09
    è¾ķ
    -0.09
    POSITIVE LOGITS
     ghost
    0.15
     ghosts
    0.14
    鬼
    0.14
    Ghost
    0.12
     Ghost
    0.11
    ghost
    0.11
    èIJ¦
    0.11
     haunting
    0.11
     Spirits
    0.10
     spirits
    0.10
    Activations Density 0.045%

    No Known Activations