© 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. 4-RES-JB
    11. 426
    Prev
    Next
    INDEX
    Explanations

    phrases related to communication, such as quotes and questions

    oai_token-act-pair · gpt-3.5-turbo

    punctuation characters and formatting symbols used in the text

    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
    edo
    -0.76
     Bengal
    -0.74
     dispers
    -0.72
    omorphic
    -0.68
    ured
    -0.67
     Vaugh
    -0.67
     destro
    -0.66
    laus
    -0.65
     Afric
    -0.65
    illary
    -0.65
    POSITIVE LOGITS
    _>
    0.91
    wcsstore
    0.90
    SOURCE
    0.85
    [[
    0.84
    MORE
    0.82
    <<
    0.80
    lations
    0.80
    PER
    0.78
    HEAD
    0.74
    QUEST
    0.72
    Activations Density 0.008%

    No Known Activations