© 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. 27238
    Prev
    Next
    INDEX
    Explanations

    important part of

    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
    å°ıåŀĭ
    -0.08
    å±ŀäºİèĩªå·±
    -0.08
     fullest
    -0.08
     внеÑĪне
    -0.08
    æݾ
    -0.08
     ìŀIJì²´
    -0.07
    诸å¦Ĥ
    -0.07
    æ¡ģ
    -0.07
    åij¨
    -0.07
    tent
    -0.07
    POSITIVE LOGITS
    phas
    0.10
    æľĭåıĭ们对
    0.10
    å¾Īå¤ļæľĭåıĭ
    0.10
    udas
    0.09
    ä¸Ńåħ³
    0.09
    çĤ¹åĩ»æŁ¥çľĭ
    0.09
    å§Ĭ
    0.09
    InstanceState
    0.09
    导è´Ń
    0.09
    _seen
    0.09
    Activations Density 0.135%

    No Known Activations