© 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. 22814
    Prev
    Next
    INDEX
    Explanations

    click event

    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
    WER
    -0.10
    änd
    -0.10
    åĪ»èĭ¦
    -0.10
    /compiler
    -0.09
    chant
    -0.09
    strate
    -0.09
    好好
    -0.09
    å¸ĸ
    -0.09
    alez
    -0.09
    对èĩªå·±çļĦ
    -0.09
    POSITIVE LOGITS
    able
    0.11
    ABLE
    0.11
    outs
    0.10
    ables
    0.10
    ableView
    0.09
    家ãģ®
    0.08
    bara
    0.08
    iness
    0.08
    èµį
    0.08
    ONA
    0.08
    Activations Density 0.048%

    No Known Activations