© 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. 4-RES-JB
    11. 7809
    Prev
    Next
    INDEX
    Explanations

    words related to cooking techniques, ingredients, food items, and actions like tasting

    oai_token-act-pair · gpt-3.5-turbo

    references to cooking methods and food preparation terms

    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
    otiation
    -0.88
    hov
    -0.85
    MpServer
    -0.77
    asive
    -0.74
     unaff
    -0.73
    pheus
    -0.71
    gencies
    -0.71
    inese
    -0.71
    ansom
    -0.69
    etheless
    -0.69
    POSITIVE LOGITS
    OME
    0.95
    asted
    0.83
    eful
    0.80
    asting
    0.79
    aste
    0.79
    IND
    0.78
    âĶĢâĶĢâĶĢâĶĢ
    0.78
    ONT
    0.73
    Redditor
    0.73
    Override
    0.72
    Activations Density 0.013%

    No Known Activations