© 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. 33-TRANSCODER-HP
    7. 52505
    Prev
    Next
    INDEX
    Explanations

    fore

    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
    Negative Logits
    firm
    -0.29
    èĩĤ
    -0.26
     paraph
    -0.25
    <Option
    -0.25
    @dynamic
    -0.25
     sustain
    -0.25
    wing
    -0.24
    RIA
    -0.24
    EEK
    -0.24
    åĬ²
    -0.24
    POSITIVE LOGITS
    cis
    0.32
    cl
    0.28
    c
    0.28
    ce
    0.28
    çĴ
    0.26
    drug
    0.26
     dru
    0.26
    -ab
    0.26
    col
    0.26
    pos
    0.25
    Activations Density 0.001%

    No Known Activations