© 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 24576
    9. 8-RES_FS24576-JB
    11. 7908
    Prev
    Next
    INDEX
    Explanations

    references to geopolitical entities and their capitals

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    481
    +0.14
    0.6%
    104
    +0.10
    0.5%
    1
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1
    +0.14
    0.04
    104
    +0.10
    0.04
    642
    +0.10
    0.03
    Negative Logits
     motivational
    -0.95
     alarms
    -0.81
    Sym
    -0.81
     reviewers
    -0.81
     reviews
    -0.80
     presentations
    -0.75
    reads
    -0.75
    Consumer
    -0.75
     antidepressants
    -0.74
     anecdotal
    -0.74
    POSITIVE LOGITS
     annexed
    1.55
     territ
    1.36
     territories
    1.35
     borders
    1.31
     annexation
    1.31
     sovereignty
    1.29
     Territories
    1.28
     territory
    1.24
     annex
    1.23
     contiguous
    1.20
    Activations Density 0.614%

    No Known Activations