© 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. 20-TRANSCODER-HP
    7. 107873
    Prev
    Next
    INDEX
    Explanations

    for

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    为主è¦ģ
    -0.28
    çIJĨäºĭéķ¿
    -0.27
    marsh
    -0.26
    BCM
    -0.26
     sourcing
    -0.25
    ä¸ºçĽ®æłĩ
    -0.24
    icorn
    -0.24
     gamm
    -0.24
    uates
    -0.24
     assort
    -0.24
    POSITIVE LOGITS
     certain
    0.29
    udies
    0.29
     urlpatterns
    0.28
    dden
    0.28
    ãĤĵãģª
    0.27
    errated
    0.27
    åIJĪçIJĨçļĦ
    0.25
     collo
    0.25
    æĮĤçĿĢ
    0.24
    ḥ
    0.24
    Activations Density 4.901%

    No Known Activations