© 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. 32-TRANSCODER-HP
    7. 43048
    Prev
    Next
    INDEX
    Explanations

    data structures

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     FIRST
    -0.27
     пеÑĢвого
    -0.26
    bring
    -0.26
    天çĦ¶
    -0.26
    oho
    -0.25
     firsthand
    -0.25
    第ä¸Ģ
    -0.24
    avage
    -0.24
     mù
    -0.24
    ãģ¡ãģ¯
    -0.24
    POSITIVE LOGITS
     representing
    0.46
     referring
    0.42
     indicating
    0.39
     describing
    0.39
     storing
    0.35
     saved
    0.35
     that
    0.34
     referencing
    0.33
     waiting
    0.32
     depicting
    0.32
    Activations Density 0.015%

    No Known Activations