© 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. 30-TRANSCODER-HP
    7. 95878
    Prev
    Next
    INDEX
    Explanations

    significance and impact

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    court
    -0.28
     Amount
    -0.28
    amount
    -0.27
    umu
    -0.27
    iled
    -0.27
     plane
    -0.27
    velope
    -0.26
    lessly
    -0.26
    pty
    -0.26
    bell
    -0.25
    POSITIVE LOGITS
     history
    0.30
    éĩįè¦ģçļĦ
    0.29
    history
    0.29
     lasting
    0.29
    éĩįè¦ģ
    0.28
    ÑĢаг
    0.28
    æĺ¾èijĹ
    0.28
     important
    0.27
     distinctive
    0.26
     possible
    0.26
    Activations Density 0.885%

    No Known Activations