© 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. 20249
    Prev
    Next
    INDEX
    Explanations

    research findings and impact

    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
    æĤ«
    -0.10
    /OR
    -0.10
    åĴı
    -0.10
    ÑĥÑĩа
    -0.09
    обÑĭ
    -0.09
    ret
    -0.09
    çĪ±æĥħ
    -0.09
    åģī
    -0.08
    INCLUDED
    -0.08
    åĪĽæĸ°èĥ½åĬĽ
    -0.08
    POSITIVE LOGITS
    4
    0.09
    æ°ª
    0.09
    2
    0.09
    3
    0.08
    0
    0.08
    6
    0.08
    éŀŃ
    0.08
    å°Ĩåħ¶
    0.08
    èµ°åħ¥
    0.08
    1
    0.08
    Activations Density 0.595%

    No Known Activations