© 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. 10-RES-JB
    11. 24027
    Prev
    Next
    INDEX
    Explanations

    references to sports teams

    oai_token-act-pair · gpt-3.5-turbo

    references to "USA" and related geographical identifiers

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    rocal
    -0.65
     olds
    -0.63
     ISBN
    -0.60
    aceae
    -0.60
    merce
    -0.59
     Pledge
    -0.57
     sucker
    -0.57
     exam
    -0.57
    ultan
    -0.57
     dow
    -0.57
    POSITIVE LOGITS
     TODAY
    1.07
    Journal
    0.84
     NETWORK
    0.82
     NEWS
    0.80
    News
    0.80
     pione
    0.79
     PRESS
    0.76
     IMAGES
    0.76
    Focus
    0.73
     Images
    0.73
    Activations Density 0.054%

    No Known Activations