© 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. 18089
    Prev
    Next
    INDEX
    Explanations

    two

    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
    plement
    -0.28
    AKE
    -0.27
    uales
    -0.27
    eree
    -0.27
    eref
    -0.26
    ernals
    -0.26
     increment
    -0.25
    endency
    -0.25
     camp
    -0.25
    èle
    -0.25
    POSITIVE LOGITS
    tfoot
    0.27
    åľ°æ®µ
    0.26
     feats
    0.26
    ê¸ī
    0.26
    Ñģен
    0.25
    ±
    0.25
    åħħ
    0.25
    .crm
    0.24
     cá»Ļng
    0.24
    ears
    0.24
    Activations Density 6.849%

    No Known Activations