© 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. 6840
    Prev
    Next
    INDEX
    Explanations

    the phrase "considering that"

    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
    481
    +0.12
    0.5%
    703
    +0.11
    0.5%
    513
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    703
    +0.12
    0.03
    397
    +0.11
    0.02
    242
    +0.10
    0.02
    Negative Logits
    inis
    -0.90
    scribe
    -0.83
    inse
    -0.80
    ernal
    -0.79
    jer
    -0.78
    orem
    -0.78
    arez
    -0.73
    okes
    -0.72
    inates
    -0.72
    ework
    -0.72
    POSITIVE LOGITS
     recent
    0.85
     how
    0.84
     inexper
    0.78
     previous
    0.75
     circumstances
    0.75
     lack
    0.73
     hindsight
    0.72
     rumours
    0.72
     rumors
    0.71
     complications
    0.71
    Activations Density 0.098%

    No Known Activations