© 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. 11-RES-JB
    11. 5820
    Prev
    Next
    INDEX
    Explanations

    mentions of surprise or dismay

    oai_token-act-pair · gpt-3.5-turbo

    expressions of shock or surprise

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    •layer 11 - surprise

    No Comments

    Negative Logits
    ãģ®éŃĶ
    -0.76
    igmatic
    -0.69
    vantage
    -0.68
    MX
    -0.65
    soever
    -0.64
     teasp
    -0.63
    cano
    -0.62
    pection
    -0.62
    kas
    -0.62
    isSpecialOrderable
    -0.61
    POSITIVE LOGITS
     how
    1.29
     what
    1.05
     seeing
    1.04
     witnessing
    0.95
     hearing
    0.88
     why
    0.88
     having
    0.87
     developments
    0.79
     losing
    0.79
    how
    0.79
    Activations Density 0.199%

    No Known Activations