© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 21149
    Prev
    Next
    INDEX
    Explanations

    apt, appropriate, fitting, named

    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
    loom
    -0.10
    çĸ¾
    -0.10
    åīª
    -0.10
    æĬ¤èº«ç¬¦
    -0.09
     Fountain
    -0.09
    绩
    -0.09
    GetMethod
    -0.09
    è°Ĵ
    -0.09
    _prefs
    -0.09
    å´Ĩ
    -0.09
    POSITIVE LOGITS
    æģ°å½ĵ
    0.13
     fitting
    0.11
    åĩĨç¡®
    0.11
    æģ°
    0.11
     apt
    0.10
    åIJĪéĢĤ
    0.10
     appropriate
    0.10
     appropri
    0.10
    æĺ¯å¤ļä¹Ī
    0.09
     accurate
    0.09
    Activations Density 0.054%

    No Known Activations