© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 35-TRANSCODER-HP
    7. 38929
    Prev
    Next
    INDEX
    Explanations

    carbon

    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
     the
    -0.36
     a
    -0.31
    ,");↵
    -0.30
     Same
    -0.30
    the
    -0.29
    ,S
    -0.29
    "a
    -0.29
     SS
    -0.29
     Your
    -0.28
     Ir
    -0.28
    POSITIVE LOGITS
    -loss
    0.39
    -d
    0.38
    -abortion
    0.38
    -fluid
    0.38
    -copy
    0.37
    -drive
    0.37
    -dialog
    0.37
    -transform
    0.37
    -chair
    0.37
    -cloud
    0.36
    Activations Density 0.470%

    No Known Activations