© 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. Qwen3-4B
    5. 18-TRANSCODER-HP
    7. 88642
    Prev
    Next
    INDEX
    Explanations

    long texts

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    çij¾
    -0.27
    æ¯ıä¸Ģ天
    -0.27
    çģ¾åĮº
    -0.25
    .persistent
    -0.25
    èĹķ
    -0.24
    æ¯ı天
    -0.24
    å¢ŀéĩı
    -0.24
    _CAL
    -0.24
    stress
    -0.23
    çijĽ
    -0.23
    POSITIVE LOGITS
    å¦Ĵ
    0.26
    yet
    0.26
    nat
    0.26
    ç¨į
    0.25
    uestion
    0.25
    en
    0.25
     flakes
    0.24
    èĢĮæĿ¥
    0.24
     ExecutionContext
    0.24
     testim
    0.24
    Activations Density 12.852%

    No Known Activations