© 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. 36156
    Prev
    Next
    INDEX
    Explanations

    be followed by verbs

    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
    å¾Īå°ij
    -0.10
    ä¸įåIJ«
    -0.10
     лиÑĪÑĮ
    -0.09
    ous
    -0.09
    oftware
    -0.09
    almost
    -0.08
    çļĦç¼ĺ
    -0.08
    oming
    -0.08
     almost
    -0.08
     Moose
    -0.08
    POSITIVE LOGITS
    æľīä»»ä½ķ
    0.16
    ç°¸
    0.11
    ä¸Ģç¯ĩæĸĩç«ł
    0.11
     necessarily
    0.10
    ifter
    0.10
    ä»ħä»ħ
    0.10
    ä¸Ģåij³
    0.10
    ched
    0.10
    ä¸ĭä¸Ģ代
    0.09
    -wh
    0.09
    Activations Density 0.034%

    No Known Activations