© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 1989
    Prev
    Next
    INDEX
    Explanations

    phrases related to computer programming and technical instructions

    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
    367
    +0.14
    0.7%
    551
    +0.14
    0.6%
    718
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.14
    0.08
    367
    +0.14
    0.06
    139
    +0.09
    0.05
    Negative Logits
     Named
    -0.73
     Lawrence
    -0.72
     Reporting
    -0.68
     weaving
    -0.66
     Venture
    -0.65
     writing
    -0.65
     retiring
    -0.64
     handwritten
    -0.63
     Occup
    -0.62
     Kut
    -0.62
    POSITIVE LOGITS
     alleviate
    1.09
     suffice
    1.08
     entail
    1.06
     mitigate
    0.98
     preclude
    0.97
    ŃĶ
    0.96
     elicit
    0.96
     complicate
    0.96
     undermine
    0.95
     necess
    0.95
    Activations Density 0.171%

    No Known Activations