© 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. 4818
    Prev
    Next
    INDEX
    Explanations

    take a

    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
    æĮĩ导
    -0.10
    åģļçļĦäºĭ
    -0.10
    vana
    -0.10
    oola
    -0.09
    åħ¥åĽŃ
    -0.09
    åİ»è¿ĩ
    -0.08
     bats
    -0.08
    åİī
    -0.08
     Rou
    -0.08
    ä¸Ģ件
    -0.08
    POSITIVE LOGITS
    è¹ļ
    0.11
    åįĹéĿŀ
    0.10
    åĶĨ
    0.10
    è¡Įèµ°
    0.10
    èĭį
    0.09
    æ¤Ĵ
    0.09
    éľ¹
    0.09
    ozilla
    0.08
    emma
    0.08
    æij©æĵ¦
    0.08
    Activations Density 0.064%

    No Known Activations