© 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. 19360
    Prev
    Next
    INDEX
    Explanations

    explanation behind this

    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
    ç¦
    -0.10
    æĶ¶èİ·
    -0.09
    uger
    -0.09
    uzzi
    -0.09
    elman
    -0.09
    åĿıäºĨ
    -0.09
    ajar
    -0.09
    омеÑĢ
    -0.09
     graves
    -0.09
    æĶĴ
    -0.09
    POSITIVE LOGITS
    è¿Ļä¹Īåģļ
    0.16
    è¿Ļä¹Īå¤ļ
    0.14
    è¿Ļä¹Ī
    0.13
    è¿Ļæł·åģļ
    0.12
    éĢīæĭ©
    0.12
    åĪĿè¡·
    0.11
    å¦ĤæѤ
    0.11
    éĢĻ麼
    0.10
    è¿Ļä¹Īä¹ħ
    0.10
    è¿Ļä¹Ī大
    0.10
    Activations Density 0.079%

    No Known Activations