© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 6976
    Prev
    Next
    INDEX
    Explanations

    considering or thinking about actions

    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
    otas
    -0.09
    akis
    -0.09
    é¹ĥ
    -0.09
    æĹłçĸijæĺ¯
    -0.09
    褥
    -0.09
    رة
    -0.08
    _BUSY
    -0.08
     лÑİбим
    -0.08
    림
    -0.08
    æĹłçĸij
    -0.08
    POSITIVE LOGITS
    åĨĴéĻ©
    0.11
    订éĺħ
    0.10
    åĬłåħ¥
    0.09
    å°Ĩåħ¶
    0.09
    æĶ¾å¼ĥ
    0.09
    çªľ
    0.09
    为ä¹ĭ
    0.08
     Äijâu
    0.08
    è´Ńä¹°
    0.08
    éĢĨ
    0.08
    Activations Density 0.031%

    No Known Activations