© 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. 27-TRANSCODER-HP
    7. 56812
    Prev
    Next
    INDEX
    Explanations

    title

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_27.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.27.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    _ES
    -0.27
    atings
    -0.26
     geral
    -0.25
    å½Ĵå±ŀäºİ
    -0.25
    سÙħÙĬ
    -0.24
    رÙĪس
    -0.24
    coles
    -0.23
     reside
    -0.23
    åIJ¬è¯´è¿ĩ
    -0.23
    pec
    -0.23
    POSITIVE LOGITS
     COMM
    0.30
     ecl
    0.28
    itzer
    0.26
    çĽ¸æľº
    0.26
    æijĦåĥı头
    0.25
    åįĴ
    0.25
    III
    0.24
    éĺŁä¼į
    0.24
     IQ
    0.24
     figure
    0.24
    Activations Density 0.002%

    No Known Activations