© 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. 63864
    Prev
    Next
    INDEX
    Explanations

    Power

    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
    éĢĢ
    -0.28
    çĬ¶
    -0.27
    CCA
    -0.26
     доÑħод
    -0.25
    è¨Ģ
    -0.25
    è´Ń票
    -0.24
    UNE
    -0.24
    æŃ£å½ĵ
    -0.24
    oux
    -0.24
    ç½ļ
    -0.24
    POSITIVE LOGITS
    intl
    0.30
     McInt
    0.30
    æ¨Ĥ
    0.27
    clone
    0.26
     similarly
    0.25
    privileged
    0.25
     acting
    0.24
    Ĥ¨
    0.24
    mans
    0.24
     act
    0.23
    Activations Density 0.310%

    No Known Activations