© 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. 20-TRANSCODER-HP
    7. 62965
    Prev
    Next
    INDEX
    Explanations

    of

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ën
    -0.31
     hend
    -0.28
    æĬĺ
    -0.28
     lik
    -0.28
    TV
    -0.26
     eyeb
    -0.26
    çģ«
    -0.25
    è¿ĩåħ³
    -0.25
    æ¿Ĥ
    -0.25
    çĤ®
    -0.25
    POSITIVE LOGITS
    arkan
    0.28
    義
    0.28
    ±
    0.27
    è¿Ľåİ»
    0.26
    åĭŁéĽĨ
    0.25
    åīĤéĩı
    0.25
    æŶéĹ´ä¸º
    0.24
     consequ
    0.24
     vain
    0.24
     dor
    0.23
    Activations Density 3.070%

    No Known Activations