© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 36591
    Prev
    Next
    INDEX
    Explanations

    abbreviation or shorthand

    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
    åĿıäºĨ
    -0.10
    tap
    -0.10
    cen
    -0.09
    大æĪĺ
    -0.09
    zw
    -0.08
    -eyed
    -0.08
    olated
    -0.08
    /do
    -0.08
    æĹłæķĮ
    -0.08
    Tap
    -0.08
    POSITIVE LOGITS
    缩
    0.13
    ç®Ģ称
    0.11
    縮
    0.10
    TestFixture
    0.10
    缩çŁŃ
    0.10
    çľĭä¸įåĪ°
    0.09
    abbreviation
    0.09
    orthand
    0.09
    iations
    0.09
    çĶļèĩ³
    0.09
    Activations Density 0.053%

    No Known Activations