© 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. 17021
    Prev
    Next
    INDEX
    Explanations

    routines, structure, consistency, ritual

    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
    éħ£
    -0.10
    çŃı
    -0.09
    æ²¼
    -0.09
    ä¼ļéķ¿
    -0.09
    æīĭå¥Ĺ
    -0.08
    apore
    -0.08
    å±ķå¼Ģ
    -0.08
    éĻ©
    -0.08
    污
    -0.08
    欧ç¾İ
    -0.08
    POSITIVE LOGITS
    Routine
    0.12
     routines
    0.12
     continuity
    0.11
     routine
    0.11
    秩åºı
    0.11
     consistency
    0.11
    ä¾ĭè¡Į
    0.11
    仪å¼ı
    0.11
    ä½ľæģ¯
    0.11
    å¥Ĺè·¯
    0.10
    Activations Density 0.051%

    No Known Activations