© 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. 33-TRANSCODER-HP
    7. 54110
    Prev
    Next
    INDEX
    Explanations

    ends in "-tions"

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    zung
    -0.30
    chantment
    -0.30
    ÑĩеÑģÑĤво
    -0.28
    URATION
    -0.28
    lung
    -0.28
    ulsion
    -0.27
     Choice
    -0.27
    nung
    -0.27
    Criterion
    -0.27
     insight
    -0.27
    POSITIVE LOGITS
    ptions
    0.41
    itations
    0.41
    izations
    0.41
    olutions
    0.39
    isations
    0.38
    inations
    0.38
    CTIONS
    0.37
    fections
    0.37
    ations
    0.37
    Definitions
    0.37
    Activations Density 0.028%

    No Known Activations