© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 4-TRANSCODER-HP
    7. 183
    Prev
    Next
    INDEX
    Explanations

    ular

    np_max-act · gemini-2.0-flash

    the word "macular" in medical and ophthalmological contexts.

    oai_token-act-pair · claude-4-5-haikuTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_4.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.4.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     tact
    -0.27
     Tek
    -0.27
     treated
    -0.25
    ances
    -0.25
    quire
    -0.24
     tratt
    -0.24
    getTable
    -0.24
    Md
    -0.24
    roman
    -0.24
     shot
    -0.24
    POSITIVE LOGITS
    casting
    0.27
    aldo
    0.27
    reich
    0.27
     else
    0.26
    åħ¶ä»ĸçļĦ
    0.25
    æ¶Ī失äºĨ
    0.25
    ä¸ĢåĪĨéĴŁ
    0.25
    à¸Ħา
    0.25
    leton
    0.25
    HO
    0.25
    Activations Density 0.015%

    No Known Activations