© 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. 57589
    Prev
    Next
    INDEX
    Explanations

    past, ago, behind, under, far

    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
    æ³ĵ
    -0.10
     premise
    -0.09
    åºĶ该å¦Ĥä½ķ
    -0.09
    obe
    -0.09
    ä¹ı
    -0.09
    ãģŀ
    -0.09
    åIJ¦è®¤
    -0.09
    nym
    -0.09
    注æĦıåĪ°
    -0.09
    èħ°éĥ¨
    -0.08
    POSITIVE LOGITS
    è¿ĩåİ»äºĨ
    0.17
    è¿ĩåİ»
    0.17
     past
    0.15
    å¾Ģäºĭ
    0.15
    éģİåİ»
    0.15
    è¿ĩåİ»çļĦ
    0.14
     Past
    0.14
    past
    0.14
    Past
    0.13
    _past
    0.12
    Activations Density 0.068%

    No Known Activations