© 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. 45600
    Prev
    Next
    INDEX
    Explanations

    references to collective identity and ownership related to the concept of "our."

    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
    608
    +0.13
    0.6%
    362
    +0.11
    0.5%
    301
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.13
    0.06
    608
    +0.11
    0.06
    385
    +0.11
    0.05
    Negative Logits
    wic
    -0.75
    bender
    -0.74
    puff
    -0.73
     Levine
    -0.70
    yang
    -0.70
    FU
    -0.69
    quart
    -0.68
    ault
    -0.68
    76561
    -0.67
     coached
    -0.67
    POSITIVE LOGITS
     nation
    1.17
    selves
    1.13
     ancestors
    1.05
     own
    1.04
     shores
    1.02
     society
    1.02
     collective
    1.00
     democracy
    0.96
     beloved
    0.96
     civilization
    0.93
    Activations Density 0.088%

    No Known Activations