© 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. 1421
    Prev
    Next
    INDEX
    Explanations

    phrases related to contrasting viewpoints or perspectives

    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%
    616
    +0.11
    0.5%
    489
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    489
    +0.14
    0.04
    362
    +0.11
    0.04
    637
    +0.08
    0.03
    Negative Logits
    lessly
    -0.74
    ople
    -0.66
    ruciating
    -0.63
     someday
    -0.62
    ties
    -0.62
    lake
    -0.61
    wine
    -0.58
    beard
    -0.58
    stein
    -0.58
    bons
    -0.57
    POSITIVE LOGITS
     occasion
    1.16
     occasions
    1.12
     heels
    1.09
     contrary
    1.08
     basis
    1.07
     downside
    1.05
     periphery
    0.93
     flip
    0.87
     verge
    0.86
     occas
    0.85
    Activations Density 0.105%

    No Known Activations