© 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. 18839
    Prev
    Next
    INDEX
    Explanations

    without

    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
    è¹ij
    -0.27
    cff
    -0.25
    èĥ¼
    -0.25
    million
    -0.25
     ESL
    -0.24
    workflow
    -0.24
    bons
    -0.24
    ç°¿
    -0.24
     workflows
    -0.24
    MAL
    -0.23
    POSITIVE LOGITS
     reference
    0.40
    åıĤèĢĥ
    0.36
     loss
    0.35
    æįŁ
    0.31
    Asked
    0.31
     Reference
    0.30
     referencia
    0.29
    allas
    0.29
    ulled
    0.29
    reference
    0.29
    Activations Density 0.468%

    No Known Activations