© 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. 22736
    Prev
    Next
    INDEX
    Explanations

    mentions of different types of sports, specifically focusing strongly on hockey

    oai_token-act-pair · gpt-3.5-turbo

    references to hockey

    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
    Virgin
    -0.86
     EE
    -0.74
    scan
    -0.73
    EE
    -0.70
    ect
    -0.68
     Virgin
    -0.67
     Flores
    -0.67
    alid
    -0.66
    roma
    -0.66
    arr
    -0.65
    POSITIVE LOGITS
     hockey
    3.75
     Hockey
    2.85
     NHL
    2.17
     basketball
    2.14
     skating
    2.02
     rink
    2.00
     puck
    1.98
     soccer
    1.98
    ockey
    1.95
     baseball
    1.91
    Activations Density 0.007%

    No Known Activations