© 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 24576
    9. 8-RES_FS24576-JB
    11. 22033
    Prev
    Next
    INDEX
    Explanations

    the concept of 'foundation' or foundational elements in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered 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
    288
    +0.13
    0.6%
    506
    +0.10
    0.5%
    202
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    21
    +0.13
    0.02
    288
    +0.10
    0.02
    737
    +0.10
    0.02
    Negative Logits
    pmwiki
    -0.88
    quer
    -0.76
    quit
    -0.69
    opers
    -0.67
    essions
    -0.63
    haw
    -0.63
    cha
    -0.61
    glomer
    -0.61
    Raven
    -0.61
    zz
    -0.61
    POSITIVE LOGITS
     foundation
    1.09
     foundations
    1.01
     underpin
    0.96
     principles
    0.93
     groundwork
    0.89
    layer
    0.89
     pillars
    0.88
    stones
    0.88
     principle
    0.85
     framework
    0.85
    Activations Density 0.049%

    No Known Activations