© 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. 48113
    Prev
    Next
    INDEX
    Explanations

    of

    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
    çĥŁçģ«
    -0.11
    æĪijçľĭ
    -0.10
    fully
    -0.09
    tin
    -0.09
    OUS
    -0.09
    äºĪ以
    -0.09
    åĬłä»¥
    -0.09
    ãģıãĤĭ
    -0.09
    ä½łä¸į
    -0.09
    èĢ½
    -0.08
    POSITIVE LOGITS
    thon
    0.11
    ertas
    0.11
    ften
    0.11
    æĿĤ
    0.10
    ffset
    0.10
    icial
    0.10
    鼾
    0.10
    Ìģ
    0.10
    leans
    0.10
    ocket
    0.09
    Activations Density 0.325%

    No Known Activations