© 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. 9472
    Prev
    Next
    INDEX
    Explanations

    the word "stuff"

    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
    éĻĽ
    -0.10
    楮
    -0.08
    åħĨ
    -0.08
    rar
    -0.08
    åŃIJéĩĮ
    -0.08
    æīĵ好
    -0.08
    mie
    -0.08
    stor
    -0.08
    éľı
    -0.08
    ìį¨
    -0.08
    POSITIVE LOGITS
    åħ¶ä»ĸçļĦ
    0.10
    teen
    0.09
    iness
    0.09
    ulent
    0.09
    ively
    0.09
    ellaneous
    0.09
    ingers
    0.08
    warz
    0.08
    åĢ©
    0.08
    ategy
    0.08
    Activations Density 0.012%

    No Known Activations