© 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. 130281
    Prev
    Next
    INDEX
    Explanations

    Disjointed text snippets

    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
    æıIJéĢŁ
    -0.26
    )!
    -0.25
    eous
    -0.25
    <![
    -0.24
     Vict
    -0.23
    .ng
    -0.23
     histo
    -0.23
     Legends
    -0.23
     Compar
    -0.23
    ntag
    -0.23
    POSITIVE LOGITS
    æ²ŁéĢļ
    0.45
     communication
    0.44
     Communication
    0.37
     communications
    0.35
     iletiÅŁim
    0.34
    Communication
    0.34
    communication
    0.33
    éĢļä¿¡
    0.32
     dialogue
    0.32
     communicate
    0.32
    Activations Density 0.012%

    No Known Activations