© 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. 35-TRANSCODER-HP
    7. 39661
    Prev
    Next
    INDEX
    Explanations

    scientific/technical contexts

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    为空
    -0.28
    ials
    -0.25
    iT
    -0.24
    å±ĭéĩĮ
    -0.24
    家ä½ı
    -0.24
     hang
    -0.24
    è§ģè¿ĩ
    -0.24
     accordingly
    -0.23
     Vac
    -0.23
    æĿ¡è§Ħå®ļ
    -0.23
    POSITIVE LOGITS
    rail
    0.33
    alic
    0.32
    assic
    0.31
    zon
    0.31
    ristol
    0.31
    pton
    0.30
    atom
    0.30
    pora
    0.29
    atoes
    0.29
    assium
    0.29
    Activations Density 0.063%

    No Known Activations