© 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. 96062
    Prev
    Next
    INDEX
    Explanations

    >

    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
    enate
    -0.29
    åĽ´
    -0.29
     hypo
    -0.28
    ä¸Ģåħ±
    -0.27
    ible
    -0.26
     East
    -0.26
     Afr
    -0.25
    碰åĪ°
    -0.25
    åľį
    -0.25
     Gow
    -0.25
    POSITIVE LOGITS
    annel
    0.30
    ạt
    0.27
    å®īå¨ľ
    0.25
    reh
    0.25
    åĸĭ
    0.25
    楼ä¸ĭ
    0.25
    æŃ»
    0.24
    便ä¼ļ
    0.24
    ality
    0.24
    人æŃ»äº¡
    0.24
    Activations Density 0.104%

    No Known Activations