© 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. 22206
    Prev
    Next
    INDEX
    Explanations

    technical conjunctions/separators

    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
    ä¸Ģå¹ħ
    -0.09
     guard
    -0.09
    vern
    -0.08
    çģµ
    -0.08
    éĽ¯
    -0.08
    rier
    -0.08
    rega
    -0.08
    ä¸Ģçĵ¶
    -0.08
    å¤Ħå¤Ħ
    -0.08
    riers
    -0.07
    POSITIVE LOGITS
    åħ¶ä»ĸçļĦ
    0.17
    åħ¶å®ĥ
    0.14
    类似çļĦ
    0.12
     дÑĢÑĥгие
    0.12
    åĪ«çļĦ
    0.12
    主æµģ
    0.11
     other
    0.11
    /or
    0.11
    ãģĿãģ®ä»ĸ
    0.10
     autres
    0.10
    Activations Density 0.332%

    No Known Activations