© 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. 96306
    Prev
    Next
    INDEX
    Explanations

    uff

    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
    uously
    -0.29
    æ¨Ĭ
    -0.29
    润
    -0.28
    rend
    -0.28
    ure
    -0.27
    gu
    -0.26
    cury
    -0.26
    à¹ģà¸Ļ
    -0.26
    agog
    -0.25
    æķ°åĢ¼
    -0.25
    POSITIVE LOGITS
     exempt
    0.27
    Act
    0.26
    acting
    0.25
    subtract
    0.25
    atis
    0.25
    ativo
    0.25
    orton
    0.24
    Action
    0.24
    PTH
    0.24
     heels
    0.24
    Activations Density 0.937%

    No Known Activations