© 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. 33-TRANSCODER-HP
    7. 53851
    Prev
    Next
    INDEX
    Explanations

    legends and landmarks

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     sal
    -0.35
    é£ŀæī¬
    -0.29
     ob
    -0.29
     back
    -0.29
     ec
    -0.28
     ad
    -0.28
     jel
    -0.28
     sed
    -0.27
    Sal
    -0.27
     SAL
    -0.27
    POSITIVE LOGITS
    one
    0.32
    ONE
    0.32
     One
    0.30
    (one
    0.30
     Fellow
    0.29
     Point
    0.29
     Belt
    0.28
     Monkey
    0.28
    Presence
    0.27
     Union
    0.26
    Activations Density 0.011%

    No Known Activations