© 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. 88058
    Prev
    Next
    INDEX
    Explanations

    leaving

    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
    lette
    -0.31
    arning
    -0.28
    lettes
    -0.27
    ulia
    -0.26
    TPL
    -0.26
    æ°¸è¿ľ
    -0.26
    æľ¬
    -0.26
    åĮº
    -0.26
    æł¡
    -0.25
     componentWill
    -0.24
    POSITIVE LOGITS
    ä¸Ģå¹ħ
    0.28
     formally
    0.28
     verbally
    0.27
    kart
    0.27
    OLER
    0.26
     grop
    0.25
     CLOSE
    0.25
     brawl
    0.25
    åŃĺéĩı
    0.24
    oreal
    0.23
    Activations Density 0.015%

    No Known Activations