© 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. 1-RES-JB
    11. 9536
    Prev
    Next
    INDEX
    Explanations

    numerical patterns consisting of one or two digits followed by a comma and a number

    oai_token-act-pair · gpt-3.5-turbo

    references to specific numerical values, particularly the number 13

    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
     tremend
    -0.93
     tradem
    -0.80
    spread
    -0.79
     srf
    -0.79
    assetsadobe
    -0.77
    staking
    -0.76
     consolid
    -0.75
     stocking
    -0.74
    ¥ŀ
    -0.72
    ģ«
    -0.72
    POSITIVE LOGITS
    66
    0.97
    37
    0.93
    13
    0.92
    87
    0.92
    83
    0.92
    81
    0.92
    97
    0.90
    94
    0.89
    63
    0.89
    rd
    0.88
    Activations Density 0.033%

    No Known Activations