© 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. 10327
    Prev
    Next
    INDEX
    Explanations

    technical terms related to computers, networking, and data storage

    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
    748
    +0.10
    0.5%
    556
    +0.10
    0.5%
    195
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    748
    +0.10
    0.05
    223
    +0.10
    0.06
    357
    +0.10
    0.05
    Negative Logits
    someone
    -0.75
    ighed
    -0.74
    endon
    -0.74
    ocaust
    -0.72
    sung
    -0.70
    esson
    -0.69
    ifice
    -0.69
    Reviewed
    -0.67
     rattled
    -0.67
    uddin
    -0.66
    POSITIVE LOGITS
     types
    1.22
     kinds
    1.19
     distinct
    1.14
     phases
    1.07
     ways
    1.03
     categories
    0.99
     main
    0.96
     methods
    0.94
     overarching
    0.94
     tiers
    0.94
    Activations Density 0.235%

    No Known Activations