© 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. 56618
    Prev
    Next
    INDEX
    Explanations

    ease or difficulty

    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
    æİĤ
    -0.09
     instead
    -0.09
     Nim
    -0.09
    naz
    -0.09
    -0.09
     Johnston
    -0.08
    imat
    -0.08
     Hass
    -0.08
    åħħè£ķ
    -0.08
    诸å¦Ĥ
    -0.08
    POSITIVE LOGITS
    åĽŀäºĭ
    0.12
    ä¸ĢåĽŀäºĭ
    0.12
    æѤäºĭ
    0.11
    çļĦäºĭæĥħ
    0.10
     affair
    0.10
    è¿Ļ件äºĭ
    0.10
    å¾Ī好çļĦ
    0.10
    éĩįè¦ģçļĦ
    0.09
    ç»ĥ
    0.09
    è¿Ļ件äºĭæĥħ
    0.09
    Activations Density 0.180%

    No Known Activations