© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 52575
    Prev
    Next
    INDEX
    Explanations

    see the

    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
    çľĭçľĭ
    -0.10
    ï¸ı
    -0.10
    pled
    -0.10
    imately
    -0.10
    aster
    -0.09
    sch
    -0.09
    sons
    -0.09
    elines
    -0.09
    ingham
    -0.09
    ãĤıãĤĬ
    -0.09
    POSITIVE LOGITS
    /he
    0.12
    ä¸Ģçľ¼
    0.11
    ision
    0.11
    aled
    0.10
    /read
    0.10
    踪
    0.10
    æĩĤ
    0.09
    apore
    0.09
    kest
    0.09
    олÑĮко
    0.09
    Activations Density 0.107%

    No Known Activations