© 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. 108190
    Prev
    Next
    INDEX
    Explanations

    code and strings

    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
    arro
    -0.30
    .§
    -0.28
    -dismiss
    -0.27
     ÐłÐµÑģпÑĥблик
    -0.27
    vip
    -0.26
    eca
    -0.25
    è¿«
    -0.25
    驾
    -0.25
    (chalk
    -0.25
     ä¸ĭåįĪ
    -0.24
    POSITIVE LOGITS
    åīģ
    0.32
    亲çĪ±
    0.29
    åłĤ
    0.27
    çīĪæĿĥå½ĴåİŁ
    0.26
    imeter
    0.25
     overlapping
    0.25
    交ç»Ļ
    0.24
    PO
    0.24
    欢è¿İæĤ¨
    0.24
    ı
    0.23
    Activations Density 0.178%

    No Known Activations