© 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. 38259
    Prev
    Next
    INDEX
    Explanations

    bonus tracks, versions, footage, re-releases

    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
    欹
    -0.09
     stage
    -0.09
    reeze
    -0.09
     spread
    -0.09
     stages
    -0.08
    æĭīå¼Ģ
    -0.08
    åѤåĦ¿
    -0.08
    erval
    -0.08
    座ä½į
    -0.08
     CK
    -0.08
    POSITIVE LOGITS
    notated
    0.10
    潺
    0.10
    lore
    0.09
     Belmont
    0.09
    åīĬ
    0.08
     Schwe
    0.08
    ÙĪÙĩ
    0.08
    èį¦
    0.08
    ty
    0.08
    éré
    0.08
    Activations Density 0.032%

    No Known Activations