© 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. 19-TRANSCODER-HP
    7. 8450
    Prev
    Next
    INDEX
    Explanations

    ,

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_19.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.19.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åĵį
    -0.30
    ç»ı纬
    -0.26
    éŁ¿
    -0.25
    çĪ¶æ¯į
    -0.25
    HeaderText
    -0.25
     Seymour
    -0.24
    /Edit
    -0.24
    phones
    -0.24
    åħļä¸Ń央
    -0.24
    aroo
    -0.24
    POSITIVE LOGITS
     called
    0.40
    称为
    0.39
    æīĢè°ĵçļĦ
    0.38
    -called
    0.36
     termed
    0.35
    åı«åģļ
    0.33
    称ä¹ĭ为
    0.32
    称ä¹ĭ
    0.31
    被称为
    0.31
    called
    0.30
    Activations Density 0.438%

    No Known Activations