© 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. 61128
    Prev
    Next
    INDEX
    Explanations

    lower costs

    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
    åĶ®
    -0.09
    ATA
    -0.09
    ücken
    -0.09
    inement
    -0.09
    adem
    -0.09
     маÑĢ
    -0.09
    èĦ¾èĥĥ
    -0.09
    æĭĽçīĮ
    -0.08
    æĬĺ
    -0.08
    CDF
    -0.08
    POSITIVE LOGITS
     costs
    0.17
    æĪIJæľ¬
    0.15
     Costs
    0.15
    éĻįä½İæĪIJæľ¬
    0.13
    éĻĭ
    0.11
    _costs
    0.11
    ãĤ³ãĤ¹ãĥĪ
    0.10
     kost
    0.10
    è´¹çĶ¨
    0.10
    æĬ¢æķij
    0.10
    Activations Density 0.055%

    No Known Activations