© 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. 18640
    Prev
    Next
    INDEX
    Explanations

    revolve

    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
    éłĵ
    -0.28
    è¶³å¤ŁçļĦ
    -0.27
    æŶéĹ´ä¸º
    -0.26
    ä¸Ģ个
    -0.26
     adequate
    -0.26
     }),
    -0.25
    åĴĤ
    -0.25
    éĺ¶æ®µ
    -0.25
     Balls
    -0.24
    足以
    -0.24
    POSITIVE LOGITS
     around
    0.37
    åĽ´
    0.36
     Around
    0.34
    around
    0.33
     back
    0.30
     bak
    0.29
    Around
    0.28
    ç»ķ
    0.28
    -around
    0.28
    idar
    0.27
    Activations Density 0.020%

    No Known Activations