© 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. 107982
    Prev
    Next
    INDEX
    Explanations

    capacity

    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
    Tap
    -0.30
     Tap
    -0.29
    lick
    -0.29
    hop
    -0.28
     tap
    -0.27
    tap
    -0.27
    lich
    -0.26
    à¸Ķา
    -0.26
    agem
    -0.25
    ign
    -0.25
    POSITIVE LOGITS
    容纳
    0.47
    容éĩı
    0.40
    人æķ°
    0.39
     capacity
    0.38
    Capacity
    0.36
    size
    0.35
     size
    0.35
    capacity
    0.35
     accommod
    0.34
    åĩłä¸ªäºº
    0.34
    Activations Density 1.173%

    No Known Activations