© 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 768
    9. 8-RES_FS768-JB
    11. 697
    Prev
    Next
    INDEX
    Explanations

    contractions

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    608
    +0.13
    0.6%
    519
    +0.10
    0.5%
    55
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.13
    0.15
    519
    +0.10
    0.12
    566
    +0.10
    0.12
    Negative Logits
    sen
    -0.68
    TAIN
    -0.68
    ren
    -0.68
    ESE
    -0.67
    roit
    -0.65
    ·
    -0.64
    upon
    -0.64
    ={
    -0.64
    erers
    -0.63
    ANS
    -0.61
    POSITIVE LOGITS
     gotta
    1.50
     gonna
    1.47
     got
    1.25
     been
    1.17
     gotten
    1.13
    been
    0.99
     gone
    0.92
     Been
    0.92
     going
    0.92
     supposed
    0.86
    Activations Density 2.744%

    No Known Activations