© 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. 21597
    Prev
    Next
    INDEX
    Explanations

    next followed by time unit

    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
    -0.09
    åīįä¸ĸ
    -0.09
    nt
    -0.09
    ะ
    -0.09
    æĹĹä¸ĭ
    -0.08
    æľĢç»Ī
    -0.08
    isc
    -0.08
     дÑĢÑĥгиÑħ
    -0.08
    ä¹ĭå¤ľ
    -0.08
    ï¸ı
    -0.08
    POSITIVE LOGITS
    -generation
    0.17
    -door
    0.15
    -ever
    0.13
    -born
    0.11
     few
    0.11
     thing
    0.10
    -found
    0.10
    /current
    0.10
     generations
    0.10
    åĩłå¹´
    0.10
    Activations Density 0.028%

    No Known Activations