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 · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 1328
    Prev
    Next
    INDEX
    Explanations

    information related to financial and economic topics, including trade disputes, tax policies, and corporate ownership

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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

    Neuron Alignment
    Index
    Value
    % of L₁
    241
    +0.11
    0.5%
    497
    +0.10
    0.5%
    574
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    546
    +0.11
    0.10
    469
    +0.10
    0.06
    673
    +0.10
    0.07
    Negative Logits
    DragonMagazine
    -0.89
    utenberg
    -0.80
     plaque
    -0.71
     descriptor
    -0.69
     transformations
    -0.68
     remem
    -0.68
     vibrations
    -0.66
     antit
    -0.66
     horm
    -0.65
     deeds
    -0.64
    POSITIVE LOGITS
     Netherlands
    1.32
     Guam
    1.26
     Philippines
    1.24
     Spain
    1.23
     Mexico
    1.22
     France
    1.21
     Colombia
    1.21
     Belgium
    1.20
     Chile
    1.19
    England
    1.19
    Activations Density 1.709%

    No Known Activations