© 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. 18937
    Prev
    Next
    INDEX
    Explanations

    estimation

    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
    reak
    -0.33
    aby
    -0.31
    è½Ń
    -0.31
    REAK
    -0.29
    avra
    -0.28
    çŁŃ
    -0.27
    å½
    -0.27
     brewed
    -0.26
    olta
    -0.25
     HID
    -0.25
    POSITIVE LOGITS
     estimation
    0.32
    è¯Ħä¼°
    0.32
    _estimate
    0.30
     assessment
    0.30
     est
    0.29
    est
    0.29
    _estimator
    0.29
     estimating
    0.29
     estimate
    0.29
    Est
    0.29
    Activations Density 0.790%

    No Known Activations