© 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. 48466
    Prev
    Next
    INDEX
    Explanations

    keep secret

    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
    å±ħ
    -0.09
    人å±ħ
    -0.09
    	BIT
    -0.09
     Lar
    -0.08
    ãģįãģ¡
    -0.08
    é£
    -0.08
    èĩ³å°Ĭ
    -0.08
     Dread
    -0.08
    éĢĢ
    -0.08
    (fe
    -0.08
    POSITIVE LOGITS
    ç§ĺå¯Ĩ
    0.21
    çļĦç§ĺå¯Ĩ
    0.18
     secrets
    0.18
     secret
    0.15
    æİĸ
    0.14
    secret
    0.14
     Secret
    0.14
    ç¥ķ
    0.14
     secrecy
    0.14
     Secrets
    0.14
    Activations Density 0.061%

    No Known Activations