© 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. 12-TRANSCODER-HP
    7. 22555
    Prev
    Next
    INDEX
    Explanations

    cardinal directions

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_12.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.12.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Bren
    -0.27
    chez
    -0.26
    çªŁ
    -0.26
    rese
    -0.26
    laps
    -0.24
    ignon
    -0.24
    borg
    -0.24
    ãĤ¸ãĥ£ãĥ¼
    -0.24
    xCD
    -0.24
    gon
    -0.24
    POSITIVE LOGITS
    æľĪä¸Ń
    0.26
    å¿Ļç¢Į
    0.24
    OW
    0.24
    æŁ¥æī¾
    0.24
     referees
    0.24
     lavoro
    0.24
    庶
    0.24
     Subjects
    0.24
     prop
    0.24
    åĵĪåĪ©
    0.24
    Activations Density 0.001%

    No Known Activations