Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 11755
    Prev
    Next
    INDEX
    Explanations

    instances of outrage and political tension in news articles

    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
    87
    +0.18
    0.9%
    470
    +0.14
    0.7%
    547
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    470
    +0.18
    0.04
    547
    +0.14
    0.04
    87
    +0.11
    -0.03
    Negative Logits
    â̦]
    -0.79
     Quote
    -0.71
    Anyway
    -0.69
    )"
    -0.68
     allele
    -0.67
    SourceFile
    -0.66
    ³³³³³³³³
    -0.66
     Verse
    -0.64
    Zip
    -0.63
     ([
    -0.63
    POSITIVE LOGITS
     embattled
    0.91
     increasingly
    0.87
     controversies
    0.81
     grapple
    0.79
     grappling
    0.77
     hotly
    0.77
     fraught
    0.76
     contentious
    0.76
     lawmakers
    0.75
     embroiled
    0.75
    Activations Density 0.878%

    No Known Activations