© 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. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 16133
    Prev
    Next
    INDEX
    Explanations

    sequences of characters that might not correspond to any meaningful content in the document

    oai_token-act-pair · gpt-3.5-turbo

    specific numerical values or symbols, often related to dosages or quantities

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

    No Comments

    Negative Logits
    nesota
    -0.87
    NetMessage
    -0.83
    essing
    -0.81
    ynthesis
    -0.80
    sembly
    -0.79
    isner
    -0.78
     reconc
    -0.76
    osponsors
    -0.75
    Downloadha
    -0.73
    itte
    -0.72
    POSITIVE LOGITS
    ³
    1.04
    ¡
    0.96
    ł
    0.89
    »
    0.85
    ²
    0.84
    ãĥ¼ãĥ
    0.82
    ´
    0.82
    ãĤª
    0.79
    ãĥ³
    0.78
    light
    0.77
    Activations Density 0.004%

    No Known Activations