© 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. 10213
    Prev
    Next
    INDEX
    Explanations

    names of countries and regions

    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
    363
    +0.13
    0.6%
    278
    +0.12
    0.5%
    73
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    363
    +0.13
    0.06
    278
    +0.12
    0.06
    73
    +0.10
    0.06
    Negative Logits
     Spoiler
    -0.76
    Spoiler
    -0.73
     Lite
    -0.72
    Alert
    -0.72
     Scare
    -0.69
    LOCK
    -0.67
     WRITE
    -0.67
    ception
    -0.67
     Woo
    -0.66
     Werner
    -0.65
    POSITIVE LOGITS
     continents
    1.19
     territories
    1.13
     regions
    1.01
     feder
    1.00
     kingdoms
    0.97
     institutions
    0.96
     locality
    0.95
     genders
    0.95
     neighbourhoods
    0.94
     empires
    0.94
    Activations Density 0.413%

    No Known Activations