© 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 49152
    9. 8-RES_FS49152-JB
    11. 26724
    Prev
    Next
    INDEX
    Explanations

    mentions of significant events or figures in history

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    97
    +0.10
    0.5%
    284
    +0.10
    0.5%
    14
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    680
    +0.10
    0.03
    577
    +0.10
    0.03
    512
    +0.09
    0.03
    Negative Logits
     Parables
    -0.73
     âĵĺ
    -0.70
    verbs
    -0.63
     Extras
    -0.62
    enhagen
    -0.62
     redes
    -0.58
     advoc
    -0.56
    ":["
    -0.56
     possessions
    -0.56
     titles
    -0.55
    POSITIVE LOGITS
    000
    1.06
    040
    0.76
    500
    0.76
    400
    0.75
    600
    0.74
    045
    0.74
    800
    0.72
    070
    0.72
    048
    0.70
    ottest
    0.70
    Activations Density 0.050%

    No Known Activations