© 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. 20-TRANSCODER-HP
    7. 108322
    Prev
    Next
    INDEX
    Explanations

    Reminders and personal reflections

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    arro
    -0.32
    æī¿æİ¥
    -0.28
    pcf
    -0.27
    	cf
    -0.26
    GREE
    -0.25
     differs
    -0.25
    å·®å¼Ĥ
    -0.25
    ä¸ĭåįĬ
    -0.24
    grad
    -0.24
     cf
    -0.24
    POSITIVE LOGITS
     remind
    0.41
    æıIJéĨĴ
    0.40
     reminding
    0.40
     reminders
    0.38
    æģ¢å¤į
    0.37
     reminder
    0.37
    ä»»ä½ķæŶåĢĻ
    0.37
     reminded
    0.35
    æ¯ıå½ĵ
    0.35
    reminder
    0.34
    Activations Density 71.884%

    No Known Activations