© 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. 33-TRANSCODER-HP
    7. 53577
    Prev
    Next
    INDEX
    Explanations

    Non-English words

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    æĭ¨
    -0.26
    adro
    -0.26
     dynasty
    -0.25
    awks
    -0.25
    agrid
    -0.25
     chased
    -0.24
    edu
    -0.24
    stry
    -0.24
    _REST
    -0.24
    Teen
    -0.24
    POSITIVE LOGITS
    åĪ©
    0.29
     special
    0.29
    çī¹æ®Ĭ
    0.28
    ç»ĵ
    0.27
    çī¹åĪ«
    0.26
     ages
    0.26
    讲
    0.26
    æŀľ
    0.25
    ges
    0.25
    _special
    0.25
    Activations Density 0.014%

    No Known Activations