© 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. 20-TRANSCODER-HP
    7. 141296
    Prev
    Next
    INDEX
    Explanations

    Irregular texts

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ä¹ŁçŁ¥éģĵ
    -0.30
    çĶŁæĢģç³»ç»Ł
    -0.28
    æ²Ļ
    -0.26
    åį·
    -0.25
    ITO
    -0.25
    æ³¥
    -0.25
    logg
    -0.25
    èĩªçĦ¶çķĮ
    -0.25
     flow
    -0.25
    <V
    -0.24
    POSITIVE LOGITS
    uess
    0.28
    uliar
    0.27
    ubber
    0.26
    ewriter
    0.25
     nerves
    0.25
    sembler
    0.25
    次æķ°
    0.25
    BERT
    0.24
    uir
    0.24
    ate
    0.24
    Activations Density 4.673%

    No Known Activations