© 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. 9387
    Prev
    Next
    INDEX
    Explanations

    socioeconomic terms related to income and experience

    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
    481
    +0.14
    0.7%
    37
    +0.11
    0.5%
    732
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    711
    +0.14
    0.03
    344
    +0.11
    0.05
    168
    +0.10
    0.04
    Negative Logits
    onement
    -0.63
     Shogun
    -0.59
    ATCH
    -0.58
     Dmit
    -0.57
     Relax
    -0.57
    WARD
    -0.56
    Release
    -0.55
    Article
    -0.54
     again
    -0.54
     relieved
    -0.54
    POSITIVE LOGITS
     backgrounds
    1.11
     disabilities
    0.97
     histories
    0.91
     background
    0.89
     affinity
    0.87
     disposition
    0.86
     stature
    0.85
     knack
    0.84
     predis
    0.83
     penchant
    0.83
    Activations Density 0.637%

    No Known Activations