© 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 · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 18881
    Prev
    Next
    INDEX
    Explanations

    references to video game modes and features

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    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

    Neuron Alignment
    Index
    Value
    % of L₁
    650
    +0.11
    0.5%
    398
    +0.10
    0.5%
    767
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    398
    +0.11
    0.01
    530
    +0.10
    0.01
    767
    +0.09
    0.01
    Negative Logits
     ultras
    -0.63
    folk
    -0.62
    gling
    -0.60
    ogy
    -0.60
     initials
    -0.57
    alg
    -0.57
     dots
    -0.55
    expr
    -0.55
    dropping
    -0.55
    uple
    -0.55
    POSITIVE LOGITS
    allery
    0.77
     Controller
    0.70
     Powered
    0.69
    perture
    0.67
     Oops
    0.67
     ÃĹ
    0.66
     Content
    0.64
     UNCLASSIFIED
    0.63
     SHARES
    0.63
     Mode
    0.63
    Activations Density 0.012%

    No Known Activations