© 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. 37188
    Prev
    Next
    INDEX
    Explanations

    repetitions of the word "this."

    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
    159
    +0.14
    0.7%
    606
    +0.10
    0.5%
    393
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    159
    +0.14
    0.09
    186
    +0.10
    0.06
    606
    +0.10
    0.06
    Negative Logits
    agi
    -0.77
    ignt
    -0.77
    anamo
    -0.77
    hess
    -0.76
    RD
    -0.74
    aneers
    -0.73
    ARS
    -0.71
    ickets
    -0.70
    unk
    -0.68
    oller
    -0.68
    POSITIVE LOGITS
     week
    1.16
     weekend
    1.03
     year
    1.03
     latest
    0.99
     month
    0.97
     newest
    0.94
     century
    0.89
     morning
    0.86
     guy
    0.85
     decade
    0.85
    Activations Density 0.158%

    No Known Activations