© 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. 8500
    Prev
    Next
    INDEX
    Explanations

    descriptions of different types of machinery or equipment

    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
    479
    +0.18
    0.9%
    739
    +0.13
    0.6%
    6
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    739
    +0.18
    0.03
    479
    +0.13
    0.03
    588
    +0.13
    0.04
    Negative Logits
     toget
    -0.92
    theless
    -0.90
    berra
    -0.87
    ablo
    -0.79
    sterdam
    -0.77
    onda
    -0.76
     congress
    -0.75
    nown
    -0.74
    ajor
    -0.72
     answ
    -0.72
    POSITIVE LOGITS
     Opportun
    1.26
     Balanced
    1.23
     Scale
    1.23
     Replacement
    1.21
     Prevention
    1.21
     Enhancement
    1.20
     Ratio
    1.17
     Extensions
    1.17
     Quantity
    1.16
     Piercing
    1.16
    Activations Density 0.229%

    No Known Activations