© 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 24576
    9. 8-RES_FS24576-JB
    11. 23887
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "and" in close proximity to other words

    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
    409
    +0.09
    0.4%
    260
    +0.09
    0.4%
    127
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    409
    +0.09
    0.05
    512
    +0.09
    0.05
    462
    +0.09
    0.04
    Negative Logits
    :(
    -0.62
     compan
    -0.60
     deed
    -0.58
     Became
    -0.57
    elf
    -0.57
     Parables
    -0.57
     cous
    -0.56
     dogma
    -0.56
     proposition
    -0.56
     aisle
    -0.56
    POSITIVE LOGITS
    000
    1.65
    500
    1.15
    700
    1.14
    900
    1.10
    600
    1.08
    800
    1.08
    400
    1.03
    00
    1.02
    300
    0.99
    09
    0.97
    Activations Density 0.073%

    No Known Activations