© 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. 20974
    Prev
    Next
    INDEX
    Explanations

    metadata

    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
    nection
    -0.07
    пелÑĮ
    -0.07
    mend
    -0.07
    gå
    -0.06
    ovie
    -0.06
    ssize
    -0.06
    ฤà¸Ĺà¸ĺิà¹Į
    -0.06
    iven
    -0.06
    zten
    -0.06
    åŀĭä¼ģä¸ļ
    -0.06
    POSITIVE LOGITS
    ward
    0.07
    wards
    0.06
    رÙĪÙĨ
    0.06
    çĭĪ
    0.05
    quake
    0.05
    ç¬ĶèĢħ认为
    0.05
    uki
    0.05
     Cian
    0.05
    ersatz
    0.05
    ison
    0.05
    Activations Density 0.004%

    No Known Activations