© 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. 720
    Prev
    Next
    INDEX
    Explanations

    mentions of a specific name 'McClellan.'

    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
    659
    +0.13
    0.6%
    376
    +0.12
    0.6%
    103
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.13
    0.03
    103
    +0.12
    0.02
    376
    +0.10
    0.02
    Negative Logits
    DragonMagazine
    -0.76
    istics
    -0.71
    istical
    -0.69
     Defenders
    -0.69
    culosis
    -0.69
     Gemini
    -0.66
    ical
    -0.66
    ku
    -0.65
    chest
    -0.64
     Millenn
    -0.64
    POSITIVE LOGITS
    urdy
    1.19
    aleb
    1.16
    ourt
    1.10
    onduct
    1.08
    urry
    1.06
    apers
    0.99
    artney
    0.99
    allion
    0.98
    oll
    0.97
    arty
    0.97
    Activations Density 0.030%

    No Known Activations