© 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. 22-TRANSCODER-HP
    7. 129388
    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_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     formal
    -0.27
     internal
    -0.26
    mina
    -0.25
     pb
    -0.24
     green
    -0.23
     critical
    -0.23
     attention
    -0.23
    ungs
    -0.23
     common
    -0.23
    å·©
    -0.23
    POSITIVE LOGITS
    ,strong
    0.27
    仲
    0.26
    cade
    0.26
    edReader
    0.26
    å®ĭ
    0.26
    è¾½å®ģ
    0.25
    rips
    0.25
    ÑıÑĤиÑı
    0.24
    UTERS
    0.24
    ẩm
    0.23
    Activations Density 0.100%

    No Known Activations