© 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. 30-TRANSCODER-HP
    7. 95437
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    bose
    -0.27
     feas
    -0.27
    endum
    -0.25
    :error
    -0.25
    累累
    -0.25
    ä¸įæŃ»
    -0.24
    eway
    -0.24
    bject
    -0.24
    aversable
    -0.24
    fetchAll
    -0.23
    POSITIVE LOGITS
     shelter
    0.25
    ignKey
    0.25
     dispens
    0.24
    åįĥ
    0.24
     polarity
    0.23
     polym
    0.22
    ofile
    0.22
    çĤĬ
    0.22
    _assignment
    0.22
    ile
    0.22
    Activations Density 0.129%

    No Known Activations

    This feature has no known activations.