© 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. 10285
    Prev
    Next
    INDEX
    Explanations

    effect 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
    /her
    -0.11
    mage
    -0.10
    hip
    -0.10
    éļĻ
    -0.10
    rift
    -0.10
    ansi
    -0.10
    -fw
    -0.10
    .yahoo
    -0.10
    yard
    -0.09
    forg
    -0.09
    POSITIVE LOGITS
    iveness
    0.16
    uate
    0.14
    ively
    0.14
    uated
    0.13
    ors
    0.13
    ually
    0.12
    icient
    0.11
    uating
    0.11
    ivity
    0.11
    swith
    0.11
    Activations Density 0.133%

    No Known Activations