© 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. 10131
    Prev
    Next
    INDEX
    Explanations

    legal and bureaucratic language related to government orders and regulations

    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
    560
    +0.09
    0.4%
    31
    +0.09
    0.4%
    116
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.09
    0.07
    743
    +0.09
    0.03
    739
    +0.09
    0.02
    Negative Logits
     gigs
    -0.81
     hardcore
    -0.73
     hints
    -0.70
     Penguin
    -0.67
    uminati
    -0.66
     likes
    -0.65
     faked
    -0.65
     goodbye
    -0.64
     outp
    -0.64
     luckily
    -0.63
    POSITIVE LOGITS
    shall
    1.45
     Notwithstanding
    1.26
     hereby
    1.26
     ``(
    1.22
     shall
    1.19
    §
    1.15
     pursuant
    1.13
     subparagraph
    1.11
     SHALL
    1.09
     Regulations
    1.08
    Activations Density 0.635%

    No Known Activations