© 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. 0-RES-JB
    11. 11923
    Prev
    Next
    INDEX
    Explanations

    dates and time markers in a specific format (e.g., "11/21/2014" or "7:38 a.m.")

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "Updated" and its variants indicating changes or revisions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Lie
    -0.69
     destiny
    -0.64
     dream
    -0.61
     leading
    -0.61
     hair
    -0.60
     neg
    -0.60
     quarter
    -0.59
     nuts
    -0.59
     vice
    -0.58
     complicity
    -0.58
    POSITIVE LOGITS
    Updated
    4.05
    Posted
    2.01
     Updated
    1.96
    updated
    1.91
    Update
    1.76
    UPDATE
    1.55
    Reviewed
    1.46
    PDATED
    1.43
    Edited
    1.42
     updated
    1.32
    Activations Density 0.018%

    No Known Activations