© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 22240
    Prev
    Next
    INDEX
    Explanations

    repeated instances of the word "what" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    494
    +0.11
    0.5%
    430
    +0.10
    0.5%
    393
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    494
    +0.11
    0.04
    655
    +0.10
    0.04
    213
    +0.09
    0.03
    Negative Logits
     fixation
    -0.67
     fix
    -0.64
     fixing
    -0.62
     abiding
    -0.61
     concentrating
    -0.60
     hearing
    -0.60
     supporting
    -0.59
     inserting
    -0.59
     running
    -0.57
    enger
    -0.56
    POSITIVE LOGITS
    soever
    1.12
     happens
    1.06
     amounted
    1.03
     happened
    1.03
     transpired
    1.03
     constitutes
    0.98
     appears
    0.83
    wikipedia
    0.82
    Downloadha
    0.82
     resembles
    0.82
    Activations Density 0.061%

    No Known Activations