© 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. 659
    Prev
    Next
    INDEX
    Explanations

    the word "how"

    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
    699
    +0.10
    0.5%
    186
    +0.10
    0.4%
    192
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    699
    +0.10
    0.04
    60
    +0.10
    0.04
    192
    +0.09
    0.04
    Negative Logits
    76561
    -0.66
    advertisement
    -0.61
    UME
    -0.60
    holder
    -0.60
     Supplement
    -0.60
     Issue
    -0.59
     ........
    -0.58
    UM
    -0.58
    inian
    -0.57
    ograph
    -0.57
    POSITIVE LOGITS
     much
    1.07
    ls
    1.01
    beit
    0.98
    itzer
    0.91
     MUCH
    0.90
     messed
    0.88
     far
    0.87
     resilient
    0.87
     prevalent
    0.87
     badly
    0.86
    Activations Density 0.090%

    No Known Activations