© 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 3072
    9. 8-RES_FS3072-JB
    11. 80
    Prev
    Next
    INDEX
    Explanations

    names of specific sports players

    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
    526
    +0.14
    0.7%
    377
    +0.12
    0.6%
    421
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.14
    0.07
    421
    +0.12
    0.05
    251
    +0.11
    0.04
    Negative Logits
    TAIN
    -0.82
    ÄŁ
    -0.76
    CLASSIFIED
    -0.75
    cules
    -0.70
    cles
    -0.66
    puter
    -0.63
    ITAL
    -0.62
     quo
    -0.62
    RY
    -0.60
    riage
    -0.60
    POSITIVE LOGITS
    endale
    1.18
    ancing
    1.16
    adiator
    1.14
    itz
    1.13
    uten
    1.10
    acial
    1.06
    ossier
    1.06
    utton
    1.05
    itch
    1.04
    oster
    1.00
    Activations Density 0.547%

    No Known Activations