© 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. 9734
    Prev
    Next
    INDEX
    Explanations

    positive sentiments and statements of satisfaction

    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
    399
    +0.10
    0.5%
    718
    +0.08
    0.4%
    169
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    718
    +0.10
    0.09
    503
    +0.08
    0.03
    526
    +0.08
    -0.04
    Negative Logits
    enery
    -0.60
    iage
    -0.56
    ynthesis
    -0.56
     Unique
    -0.54
     Located
    -0.54
     Recre
    -0.53
     outdoors
    -0.52
     Dress
    -0.52
    naissance
    -0.51
     Gorge
    -0.51
    POSITIVE LOGITS
     cynicism
    0.82
     rhetorical
    0.78
     disingen
    0.74
     understatement
    0.69
     journalistic
    0.67
    .).
    0.66
     conclusions
    0.65
     prophetic
    0.65
     encour
    0.64
     admon
    0.64
    Activations Density 2.983%

    No Known Activations