© 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. 25-TRANSCODER-HP
    7. 107887
    Prev
    Next
    INDEX
    Explanations

    further

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     acknowledgement
    -0.26
    æĶ¸
    -0.25
    wrap
    -0.24
    B
    -0.23
    æĦļ
    -0.23
    -running
    -0.23
     straps
    -0.23
     acknowledgment
    -0.23
    -me
    -0.23
     Me
    -0.23
    POSITIVE LOGITS
     the
    0.38
     our
    0.36
     your
    0.32
    hani
    0.29
     both
    0.28
     their
    0.28
    imize
    0.27
    iates
    0.27
    embro
    0.26
    æĤ¨çļĦ
    0.26
    Activations Density 0.137%

    No Known Activations