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

    Neuronpedia

    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 1536
    9. 8-RES_FS1536-JB
    11. 1031
    Prev
    Next
    INDEX
    Explanations

    phrases related to threats and dangers, particularly in relation to security and safety

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    519
    +0.11
    0.5%
    551
    +0.10
    0.4%
    511
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.11
    0.07
    288
    +0.10
    0.11
    511
    +0.09
    0.07
    Negative Logits
    issance
    -1.00
    ional
    -0.78
    angles
    -0.72
    rative
    -0.70
    nice
    -0.69
    cellence
    -0.68
    union
    -0.68
    arist
    -0.67
    rine
    -0.66
    artney
    -0.65
    POSITIVE LOGITS
    crow
    1.11
     lest
    0.89
    mong
    0.87
     posed
    0.86
    lessly
    0.85
     lurking
    0.81
     endanger
    0.80
     jeopard
    0.79
     warnings
    0.77
    inventoryQuantity
    0.76
    Activations Density 2.208%

    No Known Activations