© 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 24576
    9. 8-RES_FS24576-JB
    11. 8090
    Prev
    Next
    INDEX
    Explanations

    phrases related to news headlines and 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_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
    373
    +0.40
    2.4%
    87
    +0.30
    1.9%
    314
    +0.18
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.40
    -0.06
    87
    +0.30
    -0.03
    756
    +0.18
    -0.03
    Negative Logits
     chuckle
    -0.42
     Architects
    -0.42
     defamation
    -0.42
     fixme
    -0.41
     jokes
    -0.41
     fallacy
    -0.41
     Schn
    -0.41
     earthquakes
    -0.40
     patented
    -0.40
    )?
    -0.40
    POSITIVE LOGITS
    »
    0.74
    ·
    0.70
    ¹
    0.66
    ¸
    0.64
    =~=~
    0.64
    ¯
    0.64
    ®
    0.64
    IJ
    0.64
    vice
    0.64
    ī
    0.62
    Activations Density 1.370%

    No Known Activations