© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 6392
    Prev
    Next
    INDEX
    Explanations

    information related to sports stadium construction and renovation projects

    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
    447
    +0.25
    1.2%
    103
    +0.10
    0.5%
    536
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    352
    +0.25
    0.15
    23
    +0.10
    0.13
    103
    +0.09
    0.13
    Negative Logits
     photoc
    -0.78
    etheless
    -0.77
    anwhile
    -0.74
    çīĪ
    -0.67
     fortun
    -0.67
     swept
    -0.65
     transc
    -0.64
    ãĥ¼ãĥĨãĤ£
    -0.62
    awaru
    -0.61
     mechanically
    -0.61
    POSITIVE LOGITS
    ¬
    0.96
    ¿
    0.95
    ¹
    0.90
    ı
    0.89
    £
    0.88
    º
    0.87
    Ī
    0.83
    ¼
    0.83
    Ħ¢
    0.82
    į
    0.82
    Activations Density 0.334%

    No Known Activations