© 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. 56997
    Prev
    Next
    INDEX
    Explanations

    space

    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
    ä¿ĥ
    -0.10
    packing
    -0.09
    unpack
    -0.09
    æ°´åĩĨ
    -0.08
     بÙĪ
    -0.08
    ibel
    -0.08
    æ¯ĶèµĽä¸Ń
    -0.08
    åĬŁ
    -0.08
    缸éĢļ
    -0.08
     Cata
    -0.08
    POSITIVE LOGITS
    空éĹ´
    0.21
     space
    0.21
    -space
    0.20
    space
    0.19
    空éĸĵ
    0.18
    _space
    0.16
     Space
    0.16
    (space
    0.15
    _SPACE
    0.15
    Space
    0.14
    Activations Density 0.200%

    No Known Activations