© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 138
    Prev
    Next
    INDEX
    Explanations

    instances where a choice or decision needs to be made between different options or actions

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    276
    +0.13
    0.6%
    109
    +0.11
    0.5%
    121
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    276
    +0.13
    0.07
    726
    +0.11
    0.05
    109
    +0.11
    0.04
    Negative Logits
    brance
    -0.80
    uum
    -0.71
    itement
    -0.69
    WARN
    -0.68
    issance
    -0.67
    peria
    -0.67
    oop
    -0.66
    amia
    -0.65
    super
    -0.64
    naissance
    -0.64
    POSITIVE LOGITS
     wisely
    0.96
     chooses
    0.83
     choice
    0.83
     choices
    0.82
     chosen
    0.82
     chose
    0.81
    axe
    0.80
     choosing
    0.77
     choose
    0.76
     randomly
    0.75
    Activations Density 0.603%

    No Known Activations