© 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. 1-RES-JB
    11. 5093
    Prev
    Next
    INDEX
    Explanations

    phrases indicating comparison or equivalence

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "matched" or similar forms indicating correspondence or alignment

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    gger
    -0.82
    ä¼
    -0.75
    wart
    -0.74
    odor
    -0.68
    aby
    -0.67
    Trade
    -0.67
    Cap
    -0.67
    issue
    -0.66
    pen
    -0.65
    Zen
    -0.64
    POSITIVE LOGITS
     matched
    1.13
    matched
    0.92
    ĸļ
    0.80
     matching
    0.75
    poons
    0.71
    xual
    0.69
     matches
    0.68
     recre
    0.68
     Match
    0.66
    uates
    0.65
    Activations Density 0.012%

    No Known Activations