© 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. 22-TRANSCODER-HP
    7. 87885
    Prev
    Next
    INDEX
    Explanations

    Scientific/Research publications

    np_max-act · gemini-2.0-flash

    the final word or phrase before a punctuation mark.

    oai_token-act-pair · deepseek-v3Triggered by @finnwit
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    dish
    -0.30
     LIMITED
    -0.28
    好äºĨ
    -0.27
    rong
    -0.27
    éł
    -0.26
    hill
    -0.26
    æİ¥åıĹäºĨ
    -0.26
    ienne
    -0.26
    èĩ´
    -0.26
    Links
    -0.26
    POSITIVE LOGITS
    æī³
    0.31
     further
    0.28
     physically
    0.28
     explicit
    0.27
     formal
    0.26
    eres
    0.26
     direct
    0.26
     anymore
    0.25
    UCH
    0.25
     specifics
    0.24
    Activations Density 0.005%

    No Known Activations