© 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. 39696
    Prev
    Next
    INDEX
    Explanations

    submit form action

    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
    /of
    -0.10
    骡
    -0.10
    碧
    -0.09
    orton
    -0.09
    upon
    -0.09
    èĶ½
    -0.09
    æĺŃ
    -0.09
    umbs
    -0.09
    umph
    -0.08
    åħ·
    -0.08
    POSITIVE LOGITS
    ter
    0.12
    ting
    0.12
    TING
    0.11
    lay
    0.09
    otron
    0.09
    åİķ
    0.09
    è¾Ķ
    0.09
    race
    0.08
    ochond
    0.08
    åªļ
    0.08
    Activations Density 0.027%

    No Known Activations