© 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 12288
    9. 8-RES_FS12288-JB
    11. 10849
    Prev
    Next
    INDEX
    Explanations

    mentions of changes or trends

    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
    120
    +0.12
    0.5%
    695
    +0.10
    0.5%
    719
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    650
    +0.12
    0.06
    55
    +0.10
    0.06
    458
    +0.10
    0.05
    Negative Logits
    Joined
    -0.76
    lication
    -0.71
    odied
    -0.68
    pora
    -0.67
    lees
    -0.65
    nor
    -0.65
    obook
    -0.64
    ledge
    -0.63
    issance
    -0.63
     badge
    -0.62
    POSITIVE LOGITS
     downhill
    1.04
     spir
    0.90
     unfolded
    0.85
     unfold
    0.83
     unfolding
    0.83
     unravel
    0.83
     transpired
    0.83
     escalated
    0.79
     smoothly
    0.78
    MpServer
    0.73
    Activations Density 0.474%

    No Known Activations