© 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. 30-TRANSCODER-HP
    7. 96370
    Prev
    Next
    INDEX
    Explanations

    first

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    times
    -0.35
     times
    -0.34
    æľŁåĨħ
    -0.33
    æŶéĹ´ä¸º
    -0.31
    _TIMES
    -0.31
    pun
    -0.31
    _times
    -0.31
    oden
    -0.30
    Time
    -0.30
    æľŁéĹ´
    -0.30
    POSITIVE LOGITS
     year
    0.38
     week
    0.36
    week
    0.31
     weekday
    0.31
     day
    0.30
     night
    0.30
     evening
    0.30
     weekend
    0.29
    year
    0.29
    çĻ½å¤©
    0.29
    Activations Density 0.121%

    No Known Activations