© 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 1536
    9. 8-RES_FS1536-JB
    11. 1162
    Prev
    Next
    INDEX
    Explanations

    quotes from interviews or articles

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    373
    +0.12
    0.6%
    649
    +0.12
    0.6%
    517
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.12
    -0.07
    659
    +0.12
    0.12
    701
    +0.09
    0.09
    Negative Logits
     nails
    -0.66
     gears
    -0.65
    ĵĺ
    -0.60
    WT
    -0.59
    ongyang
    -0.58
     traged
    -0.57
     sshd
    -0.57
    INC
    -0.57
     onions
    -0.56
     horn
    -0.55
    POSITIVE LOGITS
    hyde
    0.83
    ablishment
    0.67
    icably
    0.67
    alus
    0.66
    eteria
    0.66
    ucl
    0.65
     Warwick
    0.65
    enment
    0.64
    lict
    0.64
    asia
    0.63
    Activations Density 2.884%

    No Known Activations