© 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 12288
    9. 8-RES_FS12288-JB
    11. 9830
    Prev
    Next
    INDEX
    Explanations

    technical terms related to Microsoft Azure and data centers

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    288
    +0.13
    0.6%
    447
    +0.13
    0.6%
    493
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.06
    326
    +0.13
    0.04
    115
    +0.11
    0.04
    Negative Logits
    sonian
    -0.88
    oria
    -0.82
    itsch
    -0.80
    igel
    -0.79
    orie
    -0.78
    orical
    -0.76
    glers
    -0.74
    rouse
    -0.72
    bered
    -0.72
    kered
    -0.70
    POSITIVE LOGITS
     deployments
    1.22
     Linux
    1.10
     computing
    1.09
     architectures
    1.03
     servers
    1.02
     Server
    1.02
     deployment
    1.02
     server
    1.02
    Deploy
    1.01
     Computing
    1.01
    Activations Density 0.258%

    No Known Activations