© 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. 129913
    Prev
    Next
    INDEX
    Explanations

    long

    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
    odata
    -0.28
    åĢĻ
    -0.27
    ản
    -0.26
     motion
    -0.26
    ble
    -0.26
    èĭ¥å¹²
    -0.25
    榴
    -0.25
     Motion
    -0.24
    umpt
    -0.24
    alloc
    -0.24
    POSITIVE LOGITS
    -lasting
    0.32
    ings
    0.30
    -term
    0.29
    itud
    0.28
     overdue
    0.28
    spe
    0.27
    sWith
    0.27
     enough
    0.25
    æĪijå¾Ī
    0.25
    fieldset
    0.24
    Activations Density 0.214%

    No Known Activations