© 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. 107801
    Prev
    Next
    INDEX
    Explanations

    -made

    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
    使人
    -0.29
     Lair
    -0.27
    à¸ĵะ
    -0.26
     indoors
    -0.25
    çĸ¡
    -0.25
    人å±ħ
    -0.25
    GetProperty
    -0.25
    ä¸Ģå¼ł
    -0.25
     Hutch
    -0.24
    ä¸Ģ头
    -0.24
    POSITIVE LOGITS
    ény
    0.28
    å¼¹
    0.27
    Entered
    0.25
    theless
    0.24
    -than
    0.24
    ä¿¡çĶ¨åį¡
    0.24
    jin
    0.24
    è¿ĩ
    0.24
    ambi
    0.24
    Thing
    0.23
    Activations Density 0.951%

    No Known Activations