© 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. 52873
    Prev
    Next
    INDEX
    Explanations

    say "to"

    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
    Not in Any Lists

    No Comments

    Negative Logits
    aval
    -0.31
    çĽ¸å¯¹
    -0.27
    relative
    -0.27
     relative
    -0.26
     Str
    -0.26
     Fed
    -0.26
    erot
    -0.26
     wing
    -0.26
    wing
    -0.25
    atively
    -0.25
    POSITIVE LOGITS
    _cmos
    0.28
    塵
    0.26
     halftime
    0.25
     relegated
    0.25
    Attrs
    0.24
    å´Ń
    0.24
    ulence
    0.24
    UPLOAD
    0.24
    è¯¾æľ¬
    0.24
    ToMany
    0.24
    Activations Density 0.005%

    No Known Activations