© 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. 24199
    Prev
    Next
    INDEX
    Explanations

    prepositions and conjunctions that indicate relationships and conditions

    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
    363
    +0.12
    0.5%
    227
    +0.09
    0.4%
    634
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    582
    +0.12
    0.03
    227
    +0.09
    0.03
    755
    +0.09
    0.03
    Negative Logits
    earable
    -0.62
    hyde
    -0.60
    itutional
    -0.59
    SourceFile
    -0.58
    las
    -0.57
    haar
    -0.57
     Sparrow
    -0.57
    BuyableInstoreAndOnline
    -0.56
     Shutterstock
    -0.55
     Hung
    -0.55
    POSITIVE LOGITS
     thereof
    0.79
     accordingly
    0.78
     importantly
    0.75
     ensuing
    0.72
     consequently
    0.71
     thereto
    0.70
     prest
    0.69
     consequence
    0.68
     confir
    0.66
    erity
    0.65
    Activations Density 0.226%

    No Known Activations