© 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 49152
    9. 8-RES_FS49152-JB
    11. 37474
    Prev
    Next
    INDEX
    Explanations

    references to the word "hand."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    737
    +0.12
    0.5%
    718
    +0.11
    0.5%
    371
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    212
    +0.12
    0.03
    737
    +0.11
    0.03
    371
    +0.11
    0.02
    Negative Logits
    undown
    -0.84
    vell
    -0.75
     Atmosp
    -0.75
    ï¸
    -0.69
    Trend
    -0.67
     RAND
    -0.67
    actionDate
    -0.67
    COMPLE
    -0.67
     IPS
    -0.66
    eming
    -0.66
    POSITIVE LOGITS
    bag
    1.21
    ker
    1.04
    maid
    0.95
    chair
    0.92
     grasping
    0.91
    picked
    0.90
    prints
    0.89
    rail
    0.86
    bags
    0.85
    book
    0.84
    Activations Density 0.017%

    No Known Activations