© 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. 31-TRANSCODER-HP
    7. 18657
    Prev
    Next
    INDEX
    Explanations

    analysis of materials/systems

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åĬ¨
    -0.26
    踢
    -0.25
    çĶµåŃIJ
    -0.25
    èĮ¹
    -0.25
    adh
    -0.25
    aic
    -0.24
    å·¾
    -0.23
    å¼ķ
    -0.23
    ç¬Ķ
    -0.23
     dra
    -0.23
    POSITIVE LOGITS
     analyze
    0.32
     analyses
    0.31
     analyzed
    0.31
    chas
    0.28
    abouts
    0.28
    Analy
    0.27
     Analy
    0.27
     analy
    0.27
    _heap
    0.27
     analysed
    0.27
    Activations Density 1.848%

    No Known Activations