© 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. 57908
    Prev
    Next
    INDEX
    Explanations

    breakdown or deconstruct into components

    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
    çĹĽ
    -0.09
    igo
    -0.09
     quo
    -0.09
    å¥Ķæ³¢
    -0.09
    éĿ¢çĻ½
    -0.09
     PaÅĦstwo
    -0.09
    åĥļ
    -0.09
    ç»ij
    -0.08
    acent
    -0.08
    afone
    -0.08
    POSITIVE LOGITS
    æĭĨ
    0.16
    åĪĨ解
    0.14
    æŁĿ
    0.11
     breakdown
    0.10
    åīĸæŀIJ
    0.10
    _into
    0.09
    ç¼
    0.09
     dismant
    0.09
     into
    0.09
    å¼ĢæĿ¥
    0.09
    Activations Density 0.040%

    No Known Activations