© 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. 31-TRANSCODER-HP
    7. 19685
    Prev
    Next
    INDEX
    Explanations

    Technical documents

    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
     stere
    -0.26
    olding
    -0.25
     Gab
    -0.25
    oping
    -0.25
    ãĤ«ãĥ¼
    -0.24
    ç§ģç«ĭ
    -0.24
    opa
    -0.24
     ting
    -0.24
    lete
    -0.24
    åĽ½ãģ®
    -0.24
    POSITIVE LOGITS
    ÑĪел
    0.27
    untime
    0.26
    essel
    0.26
    venue
    0.26
    river
    0.25
    ses
    0.25
    @m
    0.25
     Trial
    0.25
     Regular
    0.24
    çĥ
    0.24
    Activations Density 0.001%

    No Known Activations