© 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 1536
    9. 8-RES_FS1536-JB
    11. 20
    Prev
    Next
    INDEX
    Explanations

    phrases related to laws, violations, and breaches in legal and regulatory contexts

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    208
    +0.10
    0.5%
    329
    +0.10
    0.5%
    204
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.10
    0.13
    608
    +0.10
    0.11
    204
    +0.10
    0.07
    Negative Logits
    reau
    -0.66
    say
    -0.65
    bing
    -0.64
    hin
    -0.63
    aping
    -0.63
    waukee
    -0.63
    eware
    -0.63
    rie
    -0.62
    uce
    -0.62
    zon
    -0.61
    POSITIVE LOGITS
     norms
    1.30
     expectations
    1.23
     principles
    1.17
     tenets
    1.16
     requirements
    1.15
     ideals
    1.11
     standards
    1.09
     guidelines
    1.09
     criteria
    1.07
     obligations
    1.07
    Activations Density 2.814%

    No Known Activations