© 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. 53877
    Prev
    Next
    INDEX
    Explanations

    Non-standard language snippets

    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
    ward
    -0.33
    ستر
    -0.32
    ger
    -0.30
    GER
    -0.29
    IPP
    -0.28
     ro
    -0.28
    ido
    -0.28
    credit
    -0.27
    bor
    -0.27
     North
    -0.27
    POSITIVE LOGITS
    ronics
    0.27
     Tout
    0.26
     ATK
    0.25
    __(*
    0.25
    ifice
    0.25
    yne
    0.25
    Attached
    0.25
     Butt
    0.25
    uchs
    0.24
    åıĮèĦļ
    0.24
    Activations Density 0.006%

    No Known Activations