© 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. 6842
    Prev
    Next
    INDEX
    Explanations

    sens, ethic, aesthetic

    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
    第ä¸Ģä½į
    -0.08
    edom
    -0.08
    äºij端
    -0.08
     ven
    -0.08
    ilo
    -0.08
    æľīçļĦ
    -0.08
    VEN
    -0.07
     xm
    -0.07
    -win
    -0.07
    eger
    -0.07
    POSITIVE LOGITS
    sWith
    0.11
    年代
    0.09
    INGS
    0.09
    åѦä¼ļäºĨ
    0.09
    raphics
    0.09
    åĥµå°¸
    0.09
    ait
    0.09
    /values
    0.09
    _pdata
    0.09
    大åĶIJ
    0.09
    Activations Density 0.005%

    No Known Activations