© 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. 4-RES-JB
    11. 20366
    Prev
    Next
    INDEX
    Explanations

    terms related to exceptional performance or situations

    oai_token-act-pair · gpt-3.5-turbo

    powerful descriptive adjectives and nouns related to sports and dramatic events

    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
    rology
    -0.79
    cius
    -0.79
    source
    -0.74
    wi
    -0.74
    Council
    -0.73
    rish
    -0.73
    Sources
    -0.72
    endment
    -0.72
    anamo
    -0.70
    endants
    -0.69
    POSITIVE LOGITS
     clutch
    1.01
     notch
    1.00
     comeback
    0.91
     knockout
    0.91
     slug
    0.88
     beating
    0.86
     hitter
    0.84
     stride
    0.83
     stellar
    0.83
     playoff
    0.82
    Activations Density 0.142%

    No Known Activations