© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 8781
    Prev
    Next
    INDEX
    Explanations

    historical and political references, particularly related to individuals and events

    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
    168
    +0.14
    0.6%
    55
    +0.11
    0.5%
    543
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.14
    0.05
    543
    +0.11
    0.04
    232
    +0.11
    0.04
    Negative Logits
    FOX
    -0.84
    Safety
    -0.83
    blogs
    -0.79
    tools
    -0.78
    ourcing
    -0.76
     Accuracy
    -0.76
    ratom
    -0.76
    malink
    -0.75
    Trend
    -0.73
    Wisconsin
    -0.73
    POSITIVE LOGITS
     Augustus
    1.40
     XVI
    1.33
     VIII
    1.25
     XIV
    1.20
     Tud
    1.19
     Ferdinand
    1.15
     XII
    1.13
     ibn
    1.13
     Herod
    1.12
     Napoleon
    1.12
    Activations Density 0.179%

    No Known Activations