© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 4691
    Prev
    Next
    INDEX
    Explanations

    technical terms related to game design and development

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    641
    +0.13
    0.6%
    567
    +0.12
    0.5%
    46
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    567
    +0.13
    0.04
    641
    +0.12
    0.05
    650
    +0.11
    0.04
    Negative Logits
     denial
    -0.71
    icides
    -0.69
     fearing
    -0.69
     suspicions
    -0.65
    cember
    -0.64
     worries
    -0.63
     understatement
    -0.63
     cautioned
    -0.63
     congratulations
    -0.62
     dismissing
    -0.62
    POSITIVE LOGITS
     usable
    1.02
     cohesive
    0.99
     sustainable
    0.96
     accessible
    0.92
    accessible
    0.91
     seamless
    0.90
     readable
    0.90
     believable
    0.88
     produ
    0.87
     conform
    0.86
    Activations Density 0.325%

    No Known Activations