© 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 3072
    9. 8-RES_FS3072-JB
    11. 233
    Prev
    Next
    INDEX
    Explanations

    punctuation marks

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    373
    +0.10
    0.5%
    720
    +0.10
    0.5%
    498
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    720
    +0.10
    0.11
    43
    +0.10
    0.08
    546
    +0.09
    0.09
    Negative Logits
    ORPG
    -0.64
     Balanced
    -0.59
    uggest
    -0.58
     Fired
    -0.57
     Ended
    -0.57
     Binding
    -0.55
     Explan
    -0.55
    corn
    -0.55
     Flavoring
    -0.52
     Respons
    -0.52
    POSITIVE LOGITS
     albeit
    1.19
     uh
    1.07
     alas
    1.04
     however
    0.96
     um
    0.95
     unsurprisingly
    0.84
     namely
    0.81
     respectively
    0.81
     moreover
    0.80
    albeit
    0.77
    Activations Density 0.995%

    No Known Activations