© 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. 30-TRANSCODER-HP
    7. 95898
    Prev
    Next
    INDEX
    Explanations

    mark

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    èĥ«
    -0.30
    Anywhere
    -0.27
    èıľé¸Ł
    -0.23
    vens
    -0.23
    åħij
    -0.23
    ove
    -0.23
    ieme
    -0.23
    èijij
    -0.23
    éģĹ
    -0.23
    bach
    -0.23
    POSITIVE LOGITS
     mark
    1.65
     marks
    1.65
     marking
    1.56
     marked
    1.48
    mark
    1.45
    æłĩè®°
    1.36
    marks
    1.34
    æłĩå¿Ĺ
    1.31
    marked
    1.30
     Mark
    1.27
    Activations Density 1.156%

    No Known Activations