© 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. 0-RES-JB
    11. 14997
    Prev
    Next
    INDEX
    Explanations

    mentions of prestigious awards, specifically the Nobel Prize

    oai_token-act-pair · gpt-3.5-turbo

    references to prestigious awards, specifically the Nobel Prize and the Pulitzer Prize

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     detach
    -0.76
    aways
    -0.70
    igans
    -0.69
    gment
    -0.64
    tenance
    -0.64
     camp
    -0.62
     tabs
    -0.62
     maintained
    -0.62
     defensively
    -0.61
    gments
    -0.61
    POSITIVE LOGITS
     Nobel
    4.27
     Pulitzer
    2.43
    Nob
    1.91
     Dalai
    1.29
     Prize
    1.21
     Heisman
    1.20
     Harvard
    1.13
     MacArthur
    1.12
     Oscar
    1.11
     Krugman
    1.10
    Activations Density 0.013%

    No Known Activations