© 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. 22722
    Prev
    Next
    INDEX
    Explanations

    fragile or delicate

    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
    гÑĥ
    -0.09
    å¯Į
    -0.09
    former
    -0.08
    Andre
    -0.08
     Fut
    -0.08
     Former
    -0.08
    奡
    -0.08
    otr
    -0.08
    _family
    -0.08
    ç¢Į
    -0.08
    POSITIVE LOGITS
    èĦĨå¼±
    0.27
     fragile
    0.22
    èĦĨ
    0.18
     delicate
    0.18
    èĸĦå¼±
    0.17
    å¨ĩ
    0.16
     frag
    0.14
     delic
    0.14
    æķıæĦŁ
    0.13
    å¼±
    0.12
    Activations Density 0.043%

    No Known Activations