© 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. 3902
    Prev
    Next
    INDEX
    Explanations

    chase or chasing

    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
    èĶij
    -0.09
    ament
    -0.09
    íĺĢ
    -0.08
    è®®äºĭ
    -0.08
    icient
    -0.08
    çŀ°
    -0.08
    ally
    -0.08
    Ñıз
    -0.08
    iana
    -0.08
    宵
    -0.08
    POSITIVE LOGITS
    ObjectContext
    0.10
    çģ°å°ĺ
    0.09
    esy
    0.09
    ðŁı»
    0.09
    æĻ¨æĬ¥
    0.09
     Äijuá»ķi
    0.09
    CKER
    0.09
    åħ±åIJĮä½ĵ
    0.08
    icles
    0.08
    writing
    0.08
    Activations Density 0.004%

    No Known Activations