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. 2553
    Prev
    Next
    INDEX
    Explanations

    years mentioned, especially related to sports events and statistics

    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
    146
    +0.12
    0.6%
    707
    +0.10
    0.4%
    497
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.12
    0.11
    306
    +0.10
    0.10
    689
    +0.10
    0.09
    Negative Logits
     ancest
    -0.72
    bars
    -0.71
    ube
    -0.67
    abulary
    -0.66
    ype
    -0.63
     hungry
    -0.62
    hed
    -0.61
    phy
    -0.61
    esses
    -0.59
    este
    -0.59
    POSITIVE LOGITS
    å¹
    1.24
    -'
    1.16
     onwards
    0.94
     onward
    0.80
     edition
    0.78
     census
    0.75
    election
    0.75
    â̲
    0.72
     Census
    0.71
     vintage
    0.70
    Activations Density 0.402%

    No Known Activations