© 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. 26745
    Prev
    Next
    INDEX
    Explanations

    over 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
    pery
    -0.10
    ubar
    -0.10
     gì
    -0.10
    urry
    -0.09
    osten
    -0.09
    обÑĢаÐ
    -0.09
    ouri
    -0.09
    olin
    -0.09
    fully
    -0.09
    alogy
    -0.09
    POSITIVE LOGITS
    comes
    0.14
    heard
    0.13
    whelming
    0.12
    tures
    0.12
    tones
    0.12
    lapping
    0.11
    kill
    0.11
    éļ¾åħ³
    0.11
    å¹´çļĦ
    0.11
    whel
    0.10
    Activations Density 0.107%

    No Known Activations