© 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. 30-TRANSCODER-HP
    7. 96442
    Prev
    Next
    INDEX
    Explanations

    truth

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     hydrogen
    -0.29
    ä¸įæĢķ
    -0.29
    åħĥå·¦åı³
    -0.26
    åĮħåĮħ
    -0.25
     nickel
    -0.25
    èµĦ产管çIJĨ
    -0.25
    åħ¨è¦ĨçĽĸ
    -0.24
    èĭ±éĽĦ
    -0.24
    ä¸Ģ款
    -0.24
     meta
    -0.24
    POSITIVE LOGITS
     continue
    0.29
     bite
    0.28
     blij
    0.28
    èĥĮæĻ¯
    0.27
    继ç»Ń
    0.27
    ordo
    0.27
    便
    0.27
    æŶéĹ´
    0.26
    ãĤªãĤ¹
    0.25
    agine
    0.25
    Activations Density 0.888%

    No Known Activations