© 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. 97525
    Prev
    Next
    INDEX
    Explanations

    things

    np_max-act · gemini-2.0-flash
    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
    ÑĢаÑĤ
    -0.32
    缦
    -0.28
    dle
    -0.27
    ä¸ĩè¾¾
    -0.26
    s
    -0.25
    ellt
    -0.25
    sst
    -0.25
    æĭ¢
    -0.24
    åıįé¦Ī
    -0.24
    shr
    -0.23
    POSITIVE LOGITS
    pec
    0.28
    人们çļĦ
    0.27
    ä¸Ģç«Ļ
    0.27
     peoples
    0.25
    .GetObject
    0.25
    åı¦ä¸Ģè¾¹
    0.25
    phas
    0.25
     Specs
    0.25
    people
    0.24
     Learned
    0.24
    Activations Density 0.082%

    No Known Activations