© 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. 57357
    Prev
    Next
    INDEX
    Explanations

    cost

    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
    æ¼Ĥ亮
    -0.10
    å§ĭ
    -0.09
    antz
    -0.09
    类产åĵģ
    -0.09
    丨
    -0.08
     smith
    -0.08
    æĥļ
    -0.08
    å¾Ļ
    -0.08
    legg
    -0.08
    idine
    -0.08
    POSITIVE LOGITS
    æĪIJæľ¬
    0.34
     cost
    0.31
    cost
    0.29
    çļĦæĪIJæľ¬
    0.29
     costs
    0.28
    ãĤ³ãĤ¹ãĥĪ
    0.27
    -cost
    0.26
    Cost
    0.26
    _cost
    0.26
    å»īä»·
    0.25
    Activations Density 0.178%

    No Known Activations