© 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. 20-TRANSCODER-HP
    7. 141857
    Prev
    Next
    INDEX
    Explanations

    £

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     shaved
    -0.26
    iesta
    -0.25
    aaS
    -0.25
    bras
    -0.25
    necessary
    -0.24
     unnecessarily
    -0.24
    ahoo
    -0.24
     Spurs
    -0.24
     incarcer
    -0.24
    bara
    -0.23
    POSITIVE LOGITS
    çĻ¾
    0.30
    atile
    0.28
     flag
    0.27
    pk
    0.26
     MIL
    0.26
     Lesser
    0.25
    <ID
    0.25
     Mull
    0.25
    UPI
    0.25
    éĬ®
    0.25
    Activations Density 5.146%

    No Known Activations