© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 3-RES-JB
    11. 593
    Prev
    Next
    INDEX
    Explanations

    phrases related to sports statistics and game outcomes, particularly emphasizing individual player performance

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "at" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    FTWARE
    -0.68
    hops
    -0.67
    cro
    -0.66
    orah
    -0.64
    alpha
    -0.64
    Russ
    -0.63
    quartered
    -0.63
    rats
    -0.63
    ships
    -0.62
    enegger
    -0.62
    POSITIVE LOGITS
     halftime
    1.20
     least
    1.16
     fullback
    0.97
     midfield
    0.96
    onement
    0.94
    letico
    0.88
    las
    0.86
     home
    0.85
     league
    0.83
     weekends
    0.82
    Activations Density 0.175%

    No Known Activations