© 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. 30-TRANSCODER-HP
    7. 95631
    Prev
    Next
    INDEX
    Explanations

    there

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ’ve
    -0.36
    obot
    -0.27
    à¸Ĺะ
    -0.27
    è²Įä¼¼
    -0.26
    ’T
    -0.26
    ’re
    -0.26
    ’av
    -0.26
    ’s
    -0.26
     seem
    -0.25
    就没
    -0.25
    POSITIVE LOGITS
    被
    0.35
    çļĦåŃĺåľ¨
    0.32
     exists
    0.32
    ä¹Łè¢«
    0.29
    existent
    0.28
    asting
    0.27
    exists
    0.27
    _EXISTS
    0.27
    CanBe
    0.26
     lie
    0.25
    Activations Density 0.890%

    No Known Activations