© 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. 19-TRANSCODER-HP
    7. 8712
    Prev
    Next
    INDEX
    Explanations

    Technical passages

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_19.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.19.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     neurop
    -0.29
    ley
    -0.27
    eks
    -0.26
    retch
    -0.26
    leo
    -0.26
    le
    -0.25
    Patterns
    -0.25
    )data
    -0.25
    _patterns
    -0.25
    -ground
    -0.25
    POSITIVE LOGITS
    because
    0.38
     because
    0.36
    åĽłä¸º
    0.36
    åĽłåħ¶
    0.35
    åĽłä¸ºå®ĥ
    0.32
     since
    0.31
    åĽłä¸ºä»ĸ
    0.31
    Because
    0.31
    çĶ±äºİ
    0.31
    since
    0.30
    Activations Density 0.085%

    No Known Activations