© 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. Qwen3.5-4B
    5. 15-RES-MATRYOSHKA-65K
    7. 20775
    Prev
    Next
    INDEX
    Explanations

    "but" or "However" signalling contrast

    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
    imientos
    -0.07
    rania
    -0.07
    iggins
    -0.06
    atorium
    -0.06
     convainc
    -0.06
    ocha
    -0.06
     ÙĩÙħÚĨ
    -0.06
    store
    -0.06
    azzino
    -0.06
    åłĤ
    -0.06
    POSITIVE LOGITS
    ä¸ĢæŦ
    0.06
    ìĦ¸
    0.05
     latter
    0.05
     widespread
    0.05
    ould
    0.04
     пеÑĢе
    0.04
     considerably
    0.04
     once
    0.04
    çļĦæŃ£ç¡®
    0.04
    çļĦ大éĩı
    0.04
    Activations Density 0.023%

    No Known Activations