© 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. 8969
    Prev
    Next
    INDEX
    Explanations

    dark

    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
    ç²Ĵ
    -0.28
    (push
    -0.26
    æıIJæĮ¯
    -0.25
     nâng
    -0.25
    çļĦåİĭåĬĽ
    -0.25
    ChildIndex
    -0.24
    åIJģ
    -0.24
    åįĩ
    -0.24
    erties
    -0.24
    cion
    -0.24
    POSITIVE LOGITS
    èĹıçĿĢ
    0.42
    éļIJèĹı
    0.38
     hidden
    0.38
    æ·±å¤Ħ
    0.36
     concealed
    0.35
    éļIJèĶ½
    0.35
     hiding
    0.33
     lurking
    0.33
    ç§ĺå¯Ĩ
    0.33
    èº²åľ¨
    0.32
    Activations Density 0.174%

    No Known Activations