© 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. 9879
    Prev
    Next
    INDEX
    Explanations

    narrative stories or historical accounts

    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
    756
    +0.10
    0.5%
    740
    +0.10
    0.5%
    556
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    390
    +0.10
    0.02
    622
    +0.10
    0.02
    56
    +0.09
    0.03
    Negative Logits
    erate
    -0.85
    hement
    -0.79
    emouth
    -0.74
    FML
    -0.74
    CPU
    -0.72
    offensive
    -0.70
    OPA
    -0.70
    tarian
    -0.69
    activated
    -0.69
    ensor
    -0.69
    POSITIVE LOGITS
     tales
    1.39
     tale
    1.22
     recounted
    1.16
     Tales
    1.09
     stories
    1.06
     narrated
    1.06
     anecdotes
    1.05
     narr
    1.04
     vividly
    1.01
     Stories
    1.00
    Activations Density 0.488%

    No Known Activations