© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 7608
    Prev
    Next
    INDEX
    Explanations

    conjunctions 'and' with a high activation value, indicating a high importance placed on relationships between ideas or concepts

    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
    373
    +0.12
    0.6%
    694
    +0.10
    0.5%
    590
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.12
    0.04
    0
    +0.10
    0.04
    484
    +0.10
    0.04
    Negative Logits
    ãĥ¢
    -0.76
    uria
    -0.70
    park
    -0.70
    file
    -0.63
    HQ
    -0.62
    leneck
    -0.61
     Locked
    -0.61
    wake
    -0.59
    olitan
    -0.59
    rison
    -0.59
    POSITIVE LOGITS
    etc
    1.30
     respectively
    1.01
     etc
    0.98
    whatever
    0.96
     blah
    0.93
     assorted
    0.88
     whatever
    0.84
     thence
    0.83
     oh
    0.81
     consequently
    0.80
    Activations Density 0.200%

    No Known Activations