© 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. 32-TRANSCODER-HP
    7. 43388
    Prev
    Next
    INDEX
    Explanations

    say "acter" ending

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åĥµ
    -0.30
    mal
    -0.26
    æŃ£è§Ħ
    -0.26
     Eyes
    -0.26
     Board
    -0.26
    é¹
    -0.25
     зак
    -0.24
    Haz
    -0.24
    çªĹå¸ĺ
    -0.24
    lix
    -0.24
    POSITIVE LOGITS
    atorial
    0.31
    actable
    0.29
    tower
    0.27
    ä¸īæĿ¡
    0.26
    æĵħ
    0.26
    acterial
    0.26
     Sure
    0.26
    iman
    0.25
    itors
    0.25
    irut
    0.24
    Activations Density 0.002%

    No Known Activations