© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 768
    5. 8-RES_FS768-JB
    6. 753
    Prev
    Next
    INDEX
    Explanations

    expressions indicating uncertainty or speculation

    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
    634
    +0.12
    0.5%
    431
    +0.09
    0.4%
    126
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.12
    0.17
    718
    +0.09
    0.17
    762
    +0.09
    0.19
    Negative Logits
    umbnails
    -0.77
     Globe
    -0.75
    aughters
    -0.73
    anwhile
    -0.73
     Feast
    -0.69
    cour
    -0.65
    ibaba
    -0.64
    otiation
    -0.64
     reminder
    -0.64
    cellaneous
    -0.64
    POSITIVE LOGITS
     somehow
    0.97
     misunderstood
    0.90
     nefarious
    0.90
     imminent
    0.89
     genuine
    0.81
     unfairly
    0.78
     legitimate
    0.77
     wrongdoing
    0.77
     wrong
    0.76
     worthwhile
    0.76
    Activations Density 7.111%

    No Known Activations