© 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 12288
    9. 8-RES_FS12288-JB
    11. 1325
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal and bureaucratic language

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    314
    +0.19
    1.0%
    87
    +0.14
    0.7%
    526
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    526
    +0.19
    -0.01
    362
    +0.14
    0.09
    283
    +0.10
    0.02
    Negative Logits
     Dictionary
    -0.55
     Emails
    -0.55
     summed
    -0.54
     typ
    -0.54
     anecd
    -0.53
    izens
    -0.52
     converge
    -0.51
     commenters
    -0.51
     lately
    -0.51
     correlations
    -0.51
    POSITIVE LOGITS
     someday
    0.84
    livion
    0.81
    instead
    0.81
     next
    0.75
     ASAP
    0.74
    ardless
    0.73
     vacated
    0.73
    ichever
    0.72
     foreseeable
    0.72
     eventual
    0.71
    Activations Density 1.503%

    No Known Activations