© 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. 51421
    Prev
    Next
    INDEX
    Explanations

    hammer, pound, percussion, playing

    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
    åĶ±
    -0.10
    åīĥ
    -0.09
    æĭī
    -0.09
    èĬ¥
    -0.09
    ripp
    -0.09
    å¯ĦãģĽ
    -0.08
    åķ®
    -0.08
    åIJ¸
    -0.08
    æłĿ
    -0.08
    oky
    -0.08
    POSITIVE LOGITS
    åĩ»
    0.20
    æķ²
    0.19
    éĶ¤
    0.19
     hammer
    0.17
    æį¶
    0.16
    æĵĬ
    0.16
    æīĵåĩ»
    0.16
    æ§Į
    0.14
    æ¦Ķ
    0.14
     Hammer
    0.13
    Activations Density 0.081%

    No Known Activations