© 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. 8-RES-JB
    11. 20030
    Prev
    Next
    INDEX
    Explanations

    software bugs and issues that have been fixed

    oai_token-act-pair · gpt-3.5-turbo

    references to bugs and fixes in software updates

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    SPONSORED
    -0.91
    rained
    -0.85
    é¾įåĸļ士
    -0.71
    walker
    -0.71
    soType
    -0.70
    ô
    -0.70
    whatever
    -0.69
    ¯
    -0.69
    chool
    -0.68
    isine
    -0.68
    POSITIVE LOGITS
     typo
    1.06
     bug
    0.89
     erroneous
    0.85
     Fixes
    0.80
     localization
    0.74
     incorrect
    0.74
     procedural
    0.74
     wording
    0.73
     compilation
    0.72
     graphical
    0.71
    Activations Density 0.105%

    No Known Activations