© 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. 7392
    Prev
    Next
    INDEX
    Explanations

    prepositions "in" or "all in"

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    341
    +0.10
    0.5%
    279
    +0.10
    0.5%
    752
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    341
    +0.10
    0.01
    616
    +0.10
    0.02
    123
    +0.10
    0.01
    Negative Logits
    RESULTS
    -0.64
     Cla
    -0.62
    stadt
    -0.60
     poets
    -0.60
    bos
    -0.59
    fell
    -0.58
    phies
    -0.58
     Malf
    -0.57
    gaard
    -0.57
    ALSE
    -0.57
    POSITIVE LOGITS
    clusive
    1.24
    clusively
    0.83
    CLUS
    0.76
    clus
    0.74
     vain
    0.74
    ighter
    0.73
    patient
    0.73
    effic
    0.69
    ixt
    0.65
     favor
    0.64
    Activations Density 0.039%

    No Known Activations