© 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. 18677
    Prev
    Next
    INDEX
    Explanations

    we

    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
    ords
    -0.27
     piracy
    -0.26
    sp
    -0.26
    _vectors
    -0.25
    andas
    -0.25
    cco
    -0.24
    å¸ĤæķĻèĤ²
    -0.24
    know
    -0.24
    ceipt
    -0.24
    wald
    -0.24
    POSITIVE LOGITS
     inspected
    0.34
     inspect
    0.33
     evaluate
    0.32
     employ
    0.31
     иÑģполÑĮзова
    0.31
     underline
    0.30
     investigates
    0.30
     Employ
    0.30
     leverage
    0.29
     evaluates
    0.29
    Activations Density 1.290%

    No Known Activations