© 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 EvalsExportsAPI 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. 3-RES-JB
    11. 20120
    Prev
    Next
    INDEX
    Explanations

    terms related to computer programming headers

    oai_token-act-pair · gpt-3.5-turbo

    mentions of header files in a programming context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    VK
    -0.77
    ingo
    -0.75
    eu
    -0.70
    Ba
    -0.69
    axis
    -0.68
    lex
    -0.67
    negie
    -0.66
    olves
    -0.65
    axies
    -0.64
    ces
    -0.64
    POSITIVE LOGITS
     headers
    1.07
     header
    1.04
    header
    0.88
     Header
    0.85
    dress
    0.83
    headers
    0.81
     acknowled
    0.78
    otle
    0.74
    layer
    0.71
    quartered
    0.70
    Activations Density 0.010%

    No Known Activations