© 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 12288
    5. 8-RES_FS12288-JB
    6. 1603
    Prev
    Next
    INDEX
    Explanations

    terms related to specific technological or scientific concepts, such as chemical compounds, historical innovations, and economic indicators

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    763
    +0.11
    0.5%
    115
    +0.11
    0.5%
    182
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.04
    115
    +0.11
    0.03
    365
    +0.09
    0.01
    Negative Logits
    Dialogue
    -0.74
    QUIRE
    -0.68
     anew
    -0.61
    Hope
    -0.60
    MAC
    -0.58
    odium
    -0.58
    Attack
    -0.58
    forum
    -0.57
    Ĥİ
    -0.57
     Kou
    -0.57
    POSITIVE LOGITS
    theless
    0.94
    oidal
    0.91
    ensical
    0.91
    restrial
    0.88
    ensable
    0.86
    igenous
    0.82
    cised
    0.79
    imensional
    0.78
    vasive
    0.77
    istant
    0.77
    Activations Density 0.174%

    No Known Activations