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

    Neuronpedia

    Jacobian LensNEW
    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. 10596
    Prev
    Next
    INDEX
    Explanations

    statements related to diplomatic discussions

    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
    480
    +0.45
    2.2%
    266
    +0.11
    0.5%
    481
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.45
    0.01
    720
    +0.11
    0.06
    679
    +0.10
    0.07
    Negative Logits
    isphere
    -0.63
     migr
    -0.63
    ahime
    -0.62
    industrial
    -0.60
    capacity
    -0.58
     disposable
    -0.56
     cubic
    -0.56
    inate
    -0.55
    variable
    -0.54
     seams
    -0.54
    POSITIVE LOGITS
     apologized
    1.04
     apologised
    0.93
     thanked
    0.88
     citing
    0.86
     noting
    0.84
     stating
    0.81
     tweeted
    0.80
     admits
    0.80
     apologizing
    0.79
     congratulated
    0.78
    Activations Density 0.624%

    No Known Activations