Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 18381
    Prev
    Next
    INDEX
    Explanations

    mentions of political events in Eastern Europe, specifically related to Crimea and annexation

    oai_token-act-pair · gpt-3.5-turbo

    references to the annexation of Crimea

    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
    ican
    -0.88
    ally
    -0.87
    lessly
    -0.83
    ochond
    -0.77
    ishes
    -0.76
    zee
    -0.74
    ient
    -0.73
    IZ
    -0.73
    nel
    -0.72
    usra
    -0.71
    POSITIVE LOGITS
    lest
    0.77
    ulators
    0.68
    aths
    0.67
    avorite
    0.67
    aucus
    0.63
    ulation
    0.63
    otle
    0.62
    otal
    0.62
    ulative
    0.61
    lets
    0.61
    Activations Density 0.139%

    No Known Activations