© 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. 22-TRANSCODER-HP
    7. 111530
    Prev
    Next
    INDEX
    Explanations

    exceptions and negations

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    çľĭä¸įè§ģ
    -0.37
    æľīæŶåĢĻ
    -0.28
     recognised
    -0.27
    creens
    -0.26
    umpt
    -0.26
     refers
    -0.26
    çľĭä¸įåĪ°
    -0.25
    GLE
    -0.24
     refer
    -0.24
     covered
    -0.24
    POSITIVE LOGITS
    INC
    0.29
     phys
    0.28
    WARD
    0.28
     correctness
    0.27
     \/
    0.27
    ward
    0.27
    æĦĪ
    0.26
     hors
    0.26
     haz
    0.25
    Correct
    0.25
    Activations Density 0.005%

    No Known Activations