© 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. 17067
    Prev
    Next
    INDEX
    Explanations

    enzyme suffixes like -ase or specific biological terms

    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
    黣
    -0.09
    eld
    -0.09
    onen
    -0.08
    æİĴéĺŁ
    -0.08
    -match
    -0.08
    _datasets
    -0.08
    uen
    -0.08
    íķĦ
    -0.08
    amel
    -0.08
    翼
    -0.08
    POSITIVE LOGITS
    æĻºéĢł
    0.09
    åĪ¶éĢłä¸ļ
    0.09
    åĮĸè¿Ľç¨ĭ
    0.09
    æĪĺèĥľ
    0.09
    ä¸į认è¯Ĩ
    0.09
    å¨ij
    0.09
    æijĨ
    0.09
    åĪĨ管
    0.08
    è¶ĬæĿ¥
    0.08
    éĴĬ
    0.08
    Activations Density 0.534%

    No Known Activations