© 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. 10107
    Prev
    Next
    INDEX
    Explanations

    time-related elements or events

    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
    +0.14
    0.6%
    463
    +0.13
    0.6%
    203
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    463
    +0.14
    0.03
    299
    +0.13
    0.02
    383
    +0.10
    0.02
    Negative Logits
    hops
    -0.81
    conom
    -0.75
    matter
    -0.73
    ovember
    -0.69
    uri
    -0.68
    rea
    -0.67
    essler
    -0.66
    ilst
    -0.65
    greg
    -0.63
    ja
    -0.62
    POSITIVE LOGITS
     reigning
    0.90
     Grammy
    0.83
     Pulitzer
    0.79
     record
    0.74
     Emmy
    0.73
     winner
    0.73
     anniversary
    0.73
     pinnacle
    0.72
     favourite
    0.71
    record
    0.71
    Activations Density 0.090%

    No Known Activations