© 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. 18794
    Prev
    Next
    INDEX
    Explanations

    rate of change

    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
     tells
    -0.26
    大家åĪĨ享
    -0.25
    iert
    -0.25
    ä¸įåıĺ
    -0.24
    .tar
    -0.24
    zell
    -0.24
    itious
    -0.23
    gradable
    -0.23
    asco
    -0.23
    iang
    -0.23
    POSITIVE LOGITS
     losing
    0.41
     gaining
    0.40
     making
    0.34
     acquiring
    0.34
     stepping
    0.33
    losing
    0.33
     influencing
    0.32
     starting
    0.31
     impacting
    0.31
    aining
    0.31
    Activations Density 0.274%

    No Known Activations