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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 1699
    Prev
    Next
    INDEX
    Explanations

    words related to legal cases and investigations, especially involving the Justice Department

    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
    146
    +0.14
    0.6%
    762
    +0.10
    0.5%
    615
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.14
    0.03
    708
    +0.10
    0.02
    615
    +0.10
    0.02
    Negative Logits
    vari
    -0.74
    oute
    -0.70
    hy
    -0.69
     Horus
    -0.68
    obyl
    -0.67
    Posts
    -0.66
    к
    -0.66
    Redditor
    -0.65
    OPLE
    -0.65
     Hollande
    -0.65
    POSITIVE LOGITS
     subpoen
    1.03
     subpoena
    0.98
     antitrust
    0.93
    ionage
    0.92
     Accountability
    0.89
     prosec
    0.88
     attorneys
    0.88
     indicted
    0.87
     DOJ
    0.85
     Inspector
    0.85
    Activations Density 0.043%

    No Known Activations