© 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. 53962
    Prev
    Next
    INDEX
    Explanations

    lesser public license

    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
    姥
    -0.10
    èĶ¼
    -0.09
    æĪĪ
    -0.09
    èĢģåħµ
    -0.09
    .FLAG
    -0.09
    rup
    -0.08
    atem
    -0.08
    sig
    -0.08
    åĵªéĩĮ
    -0.08
    ers
    -0.08
    POSITIVE LOGITS
    hood
    0.13
    -known
    0.13
    halb
    0.11
    ño
    0.10
    ablish
    0.10
    绨
    0.09
    æĥħæĢĢ
    0.09
    most
    0.09
    mente
    0.09
    oning
    0.09
    Activations Density 0.011%

    No Known Activations