© 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. 25-TRANSCODER-HP
    7. 108091
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    á»ĩ
    -0.31
    battle
    -0.28
    两类
    -0.25
    ä¸Ģçĵ¶
    -0.25
    åĩłå®¶
    -0.24
     sweetheart
    -0.24
    amine
    -0.23
    inium
    -0.23
    åħľ
    -0.23
    "',
    -0.23
    POSITIVE LOGITS
     Ung
    0.27
     AB
    0.26
     Python
    0.26
     Sum
    0.25
     PRO
    0.25
     prob
    0.25
     Pro
    0.25
     Re
    0.24
    /python
    0.24
    ::↵
    0.24
    Activations Density 0.038%

    No Known Activations