© 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 3072
    9. 8-RES_FS3072-JB
    11. 3059
    Prev
    Next
    INDEX
    Explanations

    references to political and social issues

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    58
    +0.13
    0.6%
    117
    +0.10
    0.5%
    522
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    522
    +0.13
    0.06
    58
    +0.10
    0.05
    117
    +0.10
    0.05
    Negative Logits
    ļéĨĴ
    -0.82
    ESE
    -0.79
    anwhile
    -0.71
    cffff
    -0.70
     Reply
    -0.68
    fleet
    -0.68
     srfAttach
    -0.67
    etheus
    -0.67
    IRT
    -0.66
    onto
    -0.66
    POSITIVE LOGITS
     technically
    1.38
     admittedly
    1.20
     occasional
    1.03
     initially
    0.98
     theoretically
    0.97
     disagree
    0.96
     somewhat
    0.96
     ostensibly
    0.96
     occasionally
    0.93
     undeniably
    0.89
    Activations Density 1.711%

    No Known Activations