Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 768
    5. 8-RES_FS768-JB
    6. 580
    Prev
    Next
    INDEX
    Explanations

    instances of the word 'give' or 'gave'

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    213
    +0.12
    0.5%
    508
    +0.10
    0.5%
    320
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    217
    +0.12
    0.09
    279
    +0.10
    0.10
    761
    +0.10
    0.10
    Negative Logits
    can
    -0.65
    vell
    -0.62
    psc
    -0.59
    pora
    -0.59
    atis
    -0.58
    alde
    -0.58
     proceeding
    -0.57
    ansas
    -0.57
    inyl
    -0.57
    ourge
    -0.55
    POSITIVE LOGITS
     refunds
    0.90
     birth
    0.87
     rise
    0.86
     condolences
    0.85
     generously
    0.85
    aways
    0.85
     us
    0.85
     insight
    0.83
     advice
    0.80
     unconditional
    0.78
    Activations Density 2.990%

    No Known Activations