© 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. 27918
    Prev
    Next
    INDEX
    Explanations

    as intended or expected

    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
    èľ¡
    -0.11
    ops
    -0.09
     kè
    -0.09
    æ¯ĶæĪij
    -0.09
    ÙĪاس
    -0.09
    GRES
    -0.09
    æĿijéķĩ
    -0.09
    åł°
    -0.08
     Primer
    -0.08
    æľºåŀĭ
    -0.08
    POSITIVE LOGITS
    è§Ħå®ļçļĦ
    0.10
    ëĮĢë¡ľ
    0.10
     intended
    0.10
    ipp
    0.10
    è¾¾æłĩ
    0.09
    ç¿»
    0.09
    åĪ°æľŁ
    0.09
    æ¤Ĵ
    0.08
    authorized
    0.08
    è§Ħå®ļ
    0.08
    Activations Density 0.077%

    No Known Activations