© 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. 2404
    Prev
    Next
    INDEX
    Explanations

    Can Den Stat

    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
     ÑĢабоÑĤа
    -0.09
    IJ
    -0.09
    åĩĮ
    -0.08
    çģŃçģ«
    -0.08
     disag
    -0.08
    ech
    -0.08
    пе
    -0.08
    çĶµåİĤ
    -0.08
    bole
    -0.07
    ç±³å°Ķ
    -0.07
    POSITIVE LOGITS
     Mondays
    0.16
     Madness
    0.16
    IMUM
    0.13
    ioned
    0.13
    ificent
    0.12
    Carthy
    0.12
    ellaneous
    0.12
    plier
    0.11
     Minute
    0.11
    ÙģÙĪض
    0.11
    Activations Density 1.596%

    No Known Activations