© 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. 41849
    Prev
    Next
    INDEX
    Explanations

    1

    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
    IW
    -0.29
     operational
    -0.28
     op
    -0.27
    iked
    -0.26
     power
    -0.26
    tees
    -0.25
    AndView
    -0.25
    lehem
    -0.25
    éĻIJæŶ
    -0.25
     emulate
    -0.24
    POSITIVE LOGITS
    plode
    0.29
    ÑĪек
    0.28
    æ¸Į
    0.25
     addTo
    0.25
    eller
    0.24
     Sac
    0.24
    çIJĨæĢ§
    0.24
    ä½ŀ
    0.24
    ché
    0.23
     pro
    0.23
    Activations Density 0.003%

    No Known Activations