© 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. 495
    Prev
    Next
    INDEX
    Explanations

    instructions or steps in a process

    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
    373
    +0.12
    0.6%
    595
    +0.11
    0.5%
    300
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    8
    +0.12
    0.06
    49
    +0.11
    0.05
    131
    +0.10
    0.04
    Negative Logits
     evolves
    -0.69
     progresses
    -0.67
    yth
    -0.64
     grows
    -0.63
     kills
    -0.62
     buildup
    -0.61
    UTC
    -0.60
     lasts
    -0.60
     creeps
    -0.58
     semblance
    -0.57
    POSITIVE LOGITS
     largeDownload
    0.80
    ItemImage
    0.73
    ãĤ¢ãĥ«
    0.69
    poon
    0.67
     consult
    0.66
    hari
    0.65
    rase
    0.65
    EStream
    0.65
    consider
    0.65
    rss
    0.65
    Activations Density 0.452%

    No Known Activations