© 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. 31-TRANSCODER-HP
    7. 18156
    Prev
    Next
    INDEX
    Explanations

    say "month"

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     phase
    -0.30
    lsen
    -0.25
    enger
    -0.24
    åįĬå²Ľ
    -0.24
     imaginary
    -0.24
     Jad
    -0.24
    å½±
    -0.24
    errer
    -0.24
    nad
    -0.24
    ç½²
    -0.24
    POSITIVE LOGITS
    irl
    0.30
     Pompe
    0.28
     Grat
    0.27
     Flight
    0.25
    br
    0.25
    èĮ¶åĽŃ
    0.24
    LOBAL
    0.23
    åĪ¶åĬ¨
    0.23
     Cheap
    0.22
    æīĭèĦļ
    0.22
    Activations Density 0.004%

    No Known Activations