© 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. 58972
    Prev
    Next
    INDEX
    Explanations

    suffixes like -ment, -ness, -hips

    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
    çļĦçĶŁåij½
    -0.10
    åı¯è§Ĩ
    -0.10
    uParam
    -0.09
    kker
    -0.09
    reek
    -0.09
    /MIT
    -0.09
    å¼Ģæ°´
    -0.08
    à¤ı
    -0.08
    读æĩĤ
    -0.08
    оÑĩно
    -0.08
    POSITIVE LOGITS
    ings
    0.25
    ments
    0.24
    ment
    0.22
    ation
    0.21
    ness
    0.18
    ities
    0.17
    ations
    0.16
    hips
    0.16
    ivity
    0.15
    MENT
    0.15
    Activations Density 0.567%

    No Known Activations