© 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. 6895
    Prev
    Next
    INDEX
    Explanations

    technical terms related to data analysis and processing, such as specific algorithms and methods

    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
    7
    +0.12
    0.5%
    374
    +0.11
    0.5%
    496
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    374
    +0.12
    0.01
    588
    +0.11
    0.01
    7
    +0.10
    0.01
    Negative Logits
    racuse
    -0.74
    ket
    -0.73
    guyen
    -0.68
    viron
    -0.67
    ufact
    -0.65
    Govern
    -0.64
    izations
    -0.63
    aturated
    -0.63
     lapse
    -0.61
     slideshow
    -0.61
    POSITIVE LOGITS
    ract
    1.41
    ors
    1.07
    ractor
    0.92
    ory
    0.86
    icular
    0.82
    racted
    0.80
    recated
    0.80
    ect
    0.79
    ures
    0.79
    ional
    0.78
    Activations Density 0.024%

    No Known Activations