© 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. 63277
    Prev
    Next
    INDEX
    Explanations

    original author or authors

    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
    èĢģ人家
    -0.12
    ingly
    -0.11
    é¾Ħ
    -0.10
    arat
    -0.09
    aroo
    -0.09
    æĭŃ
    -0.09
    ened
    -0.09
    ouse
    -0.09
    ä¸ĩä¸Ģ
    -0.09
    esh
    -0.09
    POSITIVE LOGITS
    ity
    0.25
    ITY
    0.18
    mente
    0.15
    ities
    0.13
    -born
    0.13
    izers
    0.12
    arily
    0.12
    ists
    0.12
    ités
    0.11
    itary
    0.11
    Activations Density 0.047%

    No Known Activations