© 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 49152
    9. 8-RES_FS49152-JB
    11. 26617
    Prev
    Next
    INDEX
    Explanations

    instances of the word "later."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    132
    +0.11
    0.5%
    174
    +0.09
    0.4%
    671
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    132
    +0.11
    0.02
    237
    +0.09
    0.02
    300
    +0.09
    0.02
    Negative Logits
     Pwr
    -0.79
    toe
    -0.73
    DEF
    -0.69
    KN
    -0.66
    cko
    -0.65
    waukee
    -0.65
    TO
    -0.65
    lua
    -0.64
    JV
    -0.63
    conservancy
    -0.63
    POSITIVE LOGITS
     generations
    0.92
    ally
    0.91
     iterations
    0.90
    noon
    0.90
     stages
    0.83
     versions
    0.82
    etheless
    0.81
     regretted
    0.80
    theless
    0.80
     installments
    0.76
    Activations Density 0.022%

    No Known Activations