© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 10241
    Prev
    Next
    INDEX
    Explanations

    key terms related to companies and organizations, particularly in the aerospace and technology industries

    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
    480
    +0.52
    3.1%
    266
    +0.11
    0.7%
    314
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.52
    0.12
    659
    +0.11
    0.10
    314
    +0.09
    0.01
    Negative Logits
     lightly
    -0.45
     indeed
    -0.42
     worthwhile
    -0.42
     VERS
    -0.42
     Patreon
    -0.42
     Gamergate
    -0.41
     GamerGate
    -0.41
     prompt
    -0.41
     outweigh
    -0.41
     carefully
    -0.41
    POSITIVE LOGITS
    undai
    0.62
    uania
    0.55
    odka
    0.55
    osh
    0.54
    tek
    0.54
    avia
    0.53
    ango
    0.53
    itan
    0.53
    ona
    0.53
    uria
    0.52
    Activations Density 1.091%

    No Known Activations