© 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. 32-TRANSCODER-HP
    7. 42348
    Prev
    Next
    INDEX
    Explanations

    ebra

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    -po
    -0.28
     Stranger
    -0.26
    éĺĸ
    -0.26
    -single
    -0.26
    çĻ¾ä¸ĩ
    -0.25
     bó
    -0.25
    енд
    -0.24
    èĵĿ天
    -0.24
     pá
    -0.24
     po
    -0.24
    POSITIVE LOGITS
    ORY
    0.28
    .construct
    0.27
    _cre
    0.27
    ç´§
    0.25
     sheer
    0.25
     constructed
    0.25
    æ°´åĪ©å·¥ç¨ĭ
    0.25
    ece
    0.25
     cre
    0.25
     motions
    0.24
    Activations Density 0.002%

    No Known Activations