© 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 24576
    9. 8-RES_FS24576-JB
    11. 18798
    Prev
    Next
    INDEX
    Explanations

    proper nouns like countries and specific 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_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
    125
    +0.12
    0.6%
    497
    +0.10
    0.4%
    241
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    546
    +0.12
    0.05
    673
    +0.10
    0.04
    121
    +0.09
    0.04
    Negative Logits
    DragonMagazine
    -0.84
    ocious
    -0.84
    utenberg
    -0.80
    reads
    -0.71
     transformative
    -0.70
    ovie
    -0.68
    otive
    -0.68
     remem
    -0.68
     disruptive
    -0.68
    ewitness
    -0.67
    POSITIVE LOGITS
     Netherlands
    1.40
     Luxembourg
    1.31
     Samoa
    1.29
     Belgium
    1.24
     France
    1.23
     Philippines
    1.23
     Romania
    1.21
     Switzerland
    1.21
     Chile
    1.21
     Denmark
    1.21
    Activations Density 0.133%

    No Known Activations