© 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. 1555
    Prev
    Next
    INDEX
    Explanations

    references to articles and posts

    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
    481
    +0.15
    0.7%
    551
    +0.12
    0.5%
    354
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.15
    0.06
    354
    +0.12
    0.05
    566
    +0.10
    0.05
    Negative Logits
    advertisement
    -0.80
    ions
    -0.65
     Replacement
    -0.62
    geries
    -0.62
    XXXX
    -0.61
     Xan
    -0.61
     Trident
    -0.57
    ridden
    -0.57
    ->
    -0.57
    WARE
    -0.57
    POSITIVE LOGITS
    've
    1.29
    'll
    1.11
    're
    1.11
     learned
    0.98
    'd
    0.93
     embarked
    0.92
     delve
    0.91
    eks
    0.85
     intend
    0.84
     propose
    0.84
    Activations Density 0.126%

    No Known Activations