© 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. 3-RES-JB
    11. 9519
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "ows" repeated throughout the document

    oai_token-act-pair · gpt-3.5-turbo

    instances of the substring "ows"

    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
     eminent
    -0.73
    âĸ¬
    -0.68
    ãĤ¶
    -0.67
     Nex
    -0.66
     apartheid
    -0.62
     nerve
    -0.62
     unaff
    -0.60
    ãĥ¤
    -0.60
    ographed
    -0.59
     gratification
    -0.58
    POSITIVE LOGITS
    erness
    1.09
    ows
    1.01
    idth
    0.97
    hops
    0.96
    itting
    0.94
    orld
    0.94
    paces
    0.92
    manship
    0.92
    hip
    0.91
    ynthesis
    0.90
    Activations Density 0.005%

    No Known Activations