© 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. 9
    Prev
    Next
    INDEX
    Explanations

    unlucky

    np_max-act · gemini-2.0-flash

    specialized or uncommon content words, especially technical/scientific terms, acronyms, and rare nouns/adjectives.

    oai_token-act-pair · gpt-5Triggered by @liuliujiujiu05
    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
    y
    -0.30
     e
    -0.28
     docs
    -0.27
     Mall
    -0.27
    .opensource
    -0.26
    Cc
    -0.25
    isel
    -0.25
     performance
    -0.24
    bie
    -0.24
    se
    -0.24
    POSITIVE LOGITS
    æĹ©æľŁ
    0.27
    innamon
    0.26
     sooner
    0.25
    onic
    0.25
     Ventures
    0.25
     TRY
    0.24
    å¼ĢåıijåĪ©çĶ¨
    0.24
    .instagram
    0.24
     Booking
    0.23
     getIntent
    0.23
    Activations Density 0.051%

    No Known Activations