© 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. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 14571
    Prev
    Next
    INDEX
    Explanations

    instances where someone is surprised or amazed

    oai_token-act-pair · gpt-3.5-turbo

    expressions of surprise or amazement

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    •layer 8 - surprise

    No Comments

    Negative Logits
    Ranked
    -0.68
    href
    -0.67
    alach
    -0.67
     fide
    -0.64
    topic
    -0.63
    prev
    -0.63
    enh
    -0.63
    haar
    -0.62
    chwitz
    -0.61
    eworks
    -0.60
    POSITIVE LOGITS
     aback
    1.00
    ãĤ¦ãĤ¹
    0.77
     by
    0.76
     enough
    0.73
     awake
    0.69
     surprised
    0.69
     how
    0.67
    >>\
    0.66
    052
    0.66
     unprepared
    0.65
    Activations Density 0.081%

    No Known Activations