© 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. 27130
    Prev
    Next
    INDEX
    Explanations

    Read more, continue, edit

    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
    	TokenName
    -0.13
    à¹Į
    -0.10
    åĮĹ大
    -0.10
    åĽºçĦ¶
    -0.09
    onical
    -0.09
    eger
    -0.09
    æĿ¾
    -0.09
    æĶ¾æīĭ
    -0.09
    istic
    -0.09
    çļĦåĬªåĬĽ
    -0.09
    POSITIVE LOGITS
     Ext
    0.09
    chief
    0.09
    åıĹåĪ°äºĨ
    0.09
    åħ¨åľº
    0.08
    åħ·ä½ĵæĥħåĨµ
    0.08
     Ne
    0.08
    .setup
    0.08
    Ext
    0.08
    æĥħåĨµè¿Ľè¡Į
    0.08
    step
    0.07
    Activations Density 0.029%

    No Known Activations