© 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. 33-TRANSCODER-HP
    7. 53323
    Prev
    Next
    INDEX
    Explanations

    atoms

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     crowd
    -0.28
    conversation
    -0.27
    olf
    -0.27
    éĥ½å¸Ĥ
    -0.26
    çĽ²
    -0.25
    .geo
    -0.25
    celain
    -0.25
     Crowd
    -0.25
    èµ·çłģ
    -0.24
    aser
    -0.24
    POSITIVE LOGITS
     atoms
    0.32
    åħ´å»º
    0.28
    åıªæľīä¸Ģ个
    0.28
     ends
    0.26
    ’s
    0.26
     building
    0.26
    ç»ĵæŀľæĺ¾ç¤º
    0.26
    's
    0.26
     layers
    0.25
    enic
    0.25
    Activations Density 0.027%

    No Known Activations