© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 2866
    Prev
    Next
    INDEX
    Explanations

    dates in the format of "YYYY-MM-DD"

    oai_token-act-pair · gpt-3.5-turbo

    references to the year 2016 and other years in the context of documentation or timelines

    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
     Th
    -0.75
     45
    -0.74
     blank
    -0.74
     50
    -0.71
     75
    -0.69
     36
    -0.69
     store
    -0.69
     th
    -0.68
     guard
    -0.68
     39
    -0.67
    POSITIVE LOGITS
    2017
    2.88
    2016
    2.83
    2015
    2.71
    2014
    2.66
    2018
    2.61
    2013
    2.59
    2012
    2.47
    2011
    2.33
    2010
    2.17
    2009
    2.09
    Activations Density 0.040%

    No Known Activations