© 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. 12-TRANSCODER-HP
    7. 23097
    Prev
    Next
    INDEX
    Explanations

    aj

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_12.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.12.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    irates
    -0.30
    æ§Ľ
    -0.29
    -heading
    -0.27
    itre
    -0.27
    orta
    -0.26
    icter
    -0.25
     crisp
    -0.25
     discomfort
    -0.25
    æĬ¼
    -0.24
    æĿł
    -0.24
    POSITIVE LOGITS
    _='
    0.28
    Spread
    0.25
    rollback
    0.25
    .Rollback
    0.25
    çĻĤ
    0.25
    etz
    0.24
    SIDE
    0.23
    åį¡æĭī
    0.23
     Spread
    0.23
    Final
    0.23
    Activations Density 0.004%

    No Known Activations