© 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. 18-TRANSCODER-HP
    7. 90228
    Prev
    Next
    INDEX
    Explanations

    mental

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    ~↵
    -0.29
    èĭŁ
    -0.26
    ftar
    -0.26
     &,
    -0.26
    åĽºçĦ¶
    -0.26
    lse
    -0.25
    æ¡´
    -0.25
    失信
    -0.25
    andel
    -0.25
    linger
    -0.24
    POSITIVE LOGITS
    olly
    0.28
     multip
    0.27
    onomy
    0.27
    uto
    0.27
     soo
    0.25
    olley
    0.25
    ologies
    0.25
    ology
    0.24
     mul
    0.24
    ä¹ĺ
    0.24
    Activations Density 24.064%

    No Known Activations