© 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. 24344
    Prev
    Next
    INDEX
    Explanations

    warnings and safety advisories in text

    oai_token-act-pair · gpt-4o-miniTriggered 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
    682
    +0.12
    0.6%
    285
    +0.12
    0.5%
    227
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    285
    +0.12
    0.05
    227
    +0.12
    0.03
    184
    +0.10
    0.03
    Negative Logits
    created
    -0.69
     descendants
    -0.69
     fame
    -0.68
     unveiling
    -0.65
     creator
    -0.64
     Divinity
    -0.64
     magically
    -0.63
     Created
    -0.63
     Noble
    -0.62
     victorious
    -0.61
    POSITIVE LOGITS
     beware
    1.62
     caution
    1.62
    eware
    1.33
    Avoid
    1.27
     precautions
    1.24
     advises
    1.21
     advise
    1.19
     Beware
    1.19
     precaution
    1.18
    avoid
    1.17
    Activations Density 0.866%

    No Known Activations