© 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. 34857
    Prev
    Next
    INDEX
    Explanations

    got a

    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
    ãģĦãģĨ
    -0.10
    auc
    -0.10
    aus
    -0.10
    çļ¿
    -0.09
    بد
    -0.08
    aq
    -0.08
    åĵį
    -0.08
    ened
    -0.08
    aub
    -0.08
    agine
    -0.08
    POSITIVE LOGITS
    çĦ¶æĺ¯
    0.12
    ter
    0.11
    terdam
    0.11
    ting
    0.11
    chas
    0.10
    /is
    0.10
    istical
    0.10
    ERIC
    0.10
    æĪĸå°Ĩ
    0.09
    tero
    0.09
    Activations Density 0.028%

    No Known Activations