© 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. Qwen3.5-4B
    5. 15-RES-MATRYOSHKA-65K
    7. 18187
    Prev
    Next
    INDEX
    Explanations

    after announcement

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    decoderesearch/qwen-3.5-saes/qwen-3.5-4b
    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
    v
    -0.05
    is
    -0.05
    m
    -0.05
     le
    -0.05
    è²Į
    -0.05
     E
    -0.05
    1
    -0.05
    ik
    -0.05
    te
    -0.04
    versions
    -0.04
    POSITIVE LOGITS
     ä»ĬæĹ¥
    0.08
    ouvel
    0.08
    oubl
    0.07
    reven
    0.07
     Caught
    0.07
     Capitale
    0.07
    âĿ¤ï¸ı
    0.07
    éı¡
    0.07
    odus
    0.07
    æīĢæĢĿ
    0.07
    Activations Density 0.544%

    No Known Activations