© 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. 19244
    Prev
    Next
    INDEX
    Explanations

    reverse or reversed roles

    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
    èį¼
    -0.10
    ä¿®åħ»
    -0.09
    åĵĦ
    -0.09
    çľ¯
    -0.09
    çĿĢæīĭ
    -0.09
     Copy
    -0.09
    å®ŀçī©
    -0.08
    ç¹ģæ®ĸ
    -0.08
    jsc
    -0.08
    UTO
    -0.08
    POSITIVE LOGITS
    éĢĨ转
    0.19
    åıį转
    0.18
     reversed
    0.17
    åıįè¿ĩæĿ¥
    0.17
     reversal
    0.16
     reverse
    0.15
    Reverse
    0.15
    .reverse
    0.15
     Reverse
    0.15
    reverse
    0.15
    Activations Density 0.081%

    No Known Activations