© 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. 52779
    Prev
    Next
    INDEX
    Explanations

    Assistant followed by specific words

    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
    all
    -0.09
    常æĢģ
    -0.09
    ToShow
    -0.08
    éĽħ
    -0.08
    ang
    -0.08
    æĭĵ
    -0.08
    åĿļæĮģ
    -0.08
    å½±åĵį
    -0.08
    _require
    -0.08
    å¼Ģæĭĵ
    -0.08
    POSITIVE LOGITS
    çŃĶå¤į
    0.10
    å¾ľ
    0.10
    蹬
    0.09
    isas
    0.09
    needle
    0.09
    oblin
    0.09
    èµĭèĥ½
    0.09
    å°¤
    0.09
    ype
    0.09
    iaux
    0.09
    Activations Density 0.016%

    No Known Activations