© 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. 4803
    Prev
    Next
    INDEX
    Explanations

    rep followed by ort, ud, ug, airs, ose, odata

    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
    lsx
    -0.09
    ä½ĵè´´
    -0.09
    obia
    -0.09
    ä¿£
    -0.09
    lers
    -0.09
    иде
    -0.09
    å·ħ
    -0.09
    ances
    -0.08
    eros
    -0.08
    ãģıãĤĭ
    -0.08
    POSITIVE LOGITS
    ertoire
    0.11
    oire
    0.11
    ACEMENT
    0.10
    resents
    0.10
    个人è§ĤçĤ¹
    0.10
    ublish
    0.09
    an
    0.09
    resenter
    0.09
    з
    0.09
    .hu
    0.09
    Activations Density 0.043%

    No Known Activations