© 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. 2332
    Prev
    Next
    INDEX
    Explanations

    phrases indicating repetitive or continuous actions

    oai_token-act-pair · gpt-3.5-turbo

    the repeated use of the words "keep" and "kept"

    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
     GOODMAN
    -0.70
    女
    -0.68
    Merit
    -0.67
    Interstitial
    -0.66
    odied
    -0.65
    issance
    -0.63
    ointed
    -0.63
    KY
    -0.62
    HCR
    -0.62
    NET
    -0.62
    POSITIVE LOGITS
     tabs
    0.93
     pace
    0.89
     track
    0.87
     afloat
    0.85
     secrets
    0.83
    tracks
    0.80
    trap
    0.78
     alive
    0.75
     repeating
    0.75
     vigil
    0.73
    Activations Density 0.037%

    No Known Activations