© 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. 30147
    Prev
    Next
    INDEX
    Explanations

    many or some cases/instances

    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
    tu
    -0.08
    mitt
    -0.08
    ileo
    -0.08
    æį®æĬ¥éģĵ
    -0.08
    av
    -0.08
    Multiplicity
    -0.08
    带çĿĢ
    -0.07
    amt
    -0.07
    .editor
    -0.07
    å°º
    -0.07
    POSITIVE LOGITS
    æĿ¥è¯´
    0.12
    èĢĮè¨Ģ
    0.12
    ä¸ĢåĪĩéĥ½
    0.11
    è¿Ļ个éĹ®é¢ĺ
    0.10
    æĪ¿
    0.10
    æĦıä¹īä¸Ĭ
    0.10
    -signed
    0.10
    æĿ¥è®²
    0.10
    roys
    0.10
    anax
    0.10
    Activations Density 0.094%

    No Known Activations