© 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. 20142
    Prev
    Next
    INDEX
    Explanations

    baking ingredients and instructions

    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
    640
    +0.23
    1.1%
    160
    +0.13
    0.6%
    442
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.23
    0.03
    160
    +0.13
    0.03
    390
    +0.10
    0.03
    Negative Logits
    ĸļ
    -0.83
     evolved
    -0.82
     evidenced
    -0.81
     furthe
    -0.76
     fielded
    -0.73
    ¥µ
    -0.72
     careers
    -0.70
     contested
    -0.70
     featured
    -0.69
     affiliated
    -0.69
    POSITIVE LOGITS
    Instruct
    1.47
    Ingredients
    1.44
    Optional
    1.33
    Prep
    1.20
    Salt
    1.19
    Recipe
    1.19
    Direct
    1.17
    ½
    1.17
    Heat
    1.15
    Method
    1.15
    Activations Density 0.046%

    No Known Activations