© 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. 95485
    Prev
    Next
    INDEX
    Explanations

    8

    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
    èİ«æĸ¯ç§ij
    -0.47
    ä¹Įåħĭåħ°
    -0.45
     Russian
    -0.42
    Putin
    -0.42
    èĭıèģĶ
    -0.41
    ä¿Ħç½Ĺæĸ¯
    -0.41
     Putin
    -0.40
    ä¸Ńä¿Ħ
    -0.40
    åŁºè¾ħ
    -0.40
    Russian
    -0.40
    POSITIVE LOGITS
    çĹ¿
    0.27
    éĵ¨
    0.27
    éħ±
    0.26
    Spain
    0.25
    French
    0.24
    dirs
    0.24
     ql
    0.24
     Spain
    0.24
     irre
    0.24
    (QL
    0.23
    Activations Density 0.885%

    No Known Activations