© 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. Qwen3-8B
    5. 18-RESID-BATCHTOPK-65K__L0-80
    7. 2431
    Prev
    Next
    INDEX
    Explanations

    supposedly allegedly magically somehow

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-8b-saes/saes_Qwen_Qwen3-8B_batch_top_k/resid_post_layer_18
    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
     policies
    -0.25
     adapters
    -0.25
     copp
    -0.23
    noch
    -0.22
     adapt
    -0.22
    duct
    -0.22
     tiếng
    -0.22
     retention
    -0.22
    als
    -0.21
    cepts
    -0.21
    POSITIVE LOGITS
     somehow
    0.55
     magically
    0.37
     Somehow
    0.34
    ä¼¼çļĦ
    0.34
     presumably
    0.29
     supposedly
    0.28
    好åIJĹ
    0.27
    ül
    0.25
     allegedly
    0.25
    "?
    0.24
    Activations Density 0.112%

    No Known Activations