© 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. 10701
    Prev
    Next
    INDEX
    Explanations

    technology-related terms and organizational structures in a business context

    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
    447
    +0.16
    0.7%
    679
    +0.11
    0.5%
    415
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.16
    0.11
    537
    +0.11
    0.10
    258
    +0.11
    0.10
    Negative Logits
    ãĥį
    -0.77
    BUG
    -0.74
    Hig
    -0.71
    çĶ°
    -0.69
    Ire
    -0.67
    rawdownloadcloneembedreportprint
    -0.66
    ãĥ´
    -0.66
    Ô
    -0.65
    Lear
    -0.64
    iest
    -0.62
    POSITIVE LOGITS
    hips
    1.39
    poons
    1.19
    hip
    1.14
    etting
    1.08
    paces
    1.07
    mith
    1.02
    ets
    1.00
    ynthesis
    1.00
    uits
    0.98
    pring
    0.97
    Activations Density 0.687%

    No Known Activations