© 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. 19349
    Prev
    Next
    INDEX
    Explanations

    until

    np_max-act-logits · gemini-2.0-flash
    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
    å¹´ä¹ĭä¹ħ
    -0.28
     kans
    -0.28
    åĩºçİ°äºĨ
    -0.26
    äºĨä¸Ģ
    -0.26
    å·²ç»ıæľīäºĨ
    -0.26
    çĬĢ
    -0.26
    æ²½
    -0.25
    eros
    -0.24
    getPost
    -0.24
     fleets
    -0.24
    POSITIVE LOGITS
     about
    0.39
    约
    0.38
     into
    0.35
     immediate
    0.34
    about
    0.33
     immediately
    0.33
     shortly
    0.33
    大约
    0.30
    Into
    0.29
    mö
    0.29
    Activations Density 0.118%

    No Known Activations