© 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. 31-TRANSCODER-HP
    7. 18651
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ooth
    -0.30
    év
    -0.28
    夹
    -0.28
     emanc
    -0.26
    assin
    -0.26
    ç»ıæµİæķĪçĽĬ
    -0.24
    æĪIJ为ä¸ŃåĽ½
    -0.24
    ussia
    -0.24
     cong
    -0.24
    .EventType
    -0.23
    POSITIVE LOGITS
    Unchecked
    0.28
    aires
    0.28
    -air
    0.26
    bara
    0.26
    çĶµä¿¡
    0.26
     unnamed
    0.25
    Deserializer
    0.25
    iring
    0.25
    ä¼Ļ
    0.25
    unk
    0.24
    Activations Density 0.002%

    No Known Activations

    This feature has no known activations.