© 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. 23630
    Prev
    Next
    INDEX
    Explanations

    pitching statistics in baseball, particularly related to striking out batters

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to striking out and scoring in sports

    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
     coolest
    -0.62
    cius
    -0.61
    Operation
    -0.60
     OPER
    -0.59
     Ore
    -0.59
     âĹ
    -0.58
     Scotia
    -0.58
     CLR
    -0.58
    Movie
    -0.57
    clusively
    -0.57
    POSITIVE LOGITS
    played
    0.92
    worked
    0.88
    stretched
    0.87
     waivers
    0.86
    whel
    0.83
    scoring
    0.81
    whelming
    0.80
    reb
    0.75
    luster
    0.74
    weak
    0.74
    Activations Density 0.091%

    No Known Activations