© 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. 31-TRANSCODER-HP
    7. 19350
    Prev
    Next
    INDEX
    Explanations

    sub

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    å¾Ģ
    -0.26
    éĩijå±±
    -0.26
    constants
    -0.25
    iscal
    -0.25
    ãĤ·ãĤ§
    -0.24
    oot
    -0.24
    ATORY
    -0.24
    ching
    -0.24
    otine
    -0.24
    ading
    -0.24
    POSITIVE LOGITS
    éĹ®æĪij
    0.31
    .optimize
    0.28
    ä¸Ģæł¹
    0.27
     optimized
    0.27
    ifu
    0.26
    è¿ij代
    0.26
    inn
    0.26
    optimized
    0.25
     optimize
    0.25
    arus
    0.25
    Activations Density 0.264%

    No Known Activations