© 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. 25-TRANSCODER-HP
    7. 108150
    Prev
    Next
    INDEX
    Explanations

    Technical or mixed content

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Lean
    -0.27
    ç»ıæµİ社ä¼ļ
    -0.26
    [d
    -0.26
    ($__
    -0.25
     ...↵↵
    -0.25
     hypo
    -0.25
    eco
    -0.25
    #↵↵
    -0.24
    ï¼Į↵
    -0.24
    ï¼Į并
    -0.24
    POSITIVE LOGITS
    reet
    0.30
    åĪ¹
    0.29
    walker
    0.28
    奢
    0.26
    è¡£
    0.26
    eryl
    0.26
    ç®±
    0.25
    çŃı
    0.25
    Eta
    0.25
    cloth
    0.25
    Activations Density 0.908%

    No Known Activations