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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 8173
    Prev
    Next
    INDEX
    Explanations

    references to space exploration, specifically regarding Mars

    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
    362
    +0.12
    0.5%
    481
    +0.10
    0.4%
    56
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.12
    0.07
    637
    +0.10
    0.04
    491
    +0.10
    0.03
    Negative Logits
    employ
    -0.85
    acters
    -0.83
    verbal
    -0.82
    cci
    -0.79
    Assistant
    -0.78
    boxing
    -0.78
    law
    -0.77
    OWN
    -0.76
    videos
    -0.75
    ãĥĥãĥī
    -0.73
    POSITIVE LOGITS
     caves
    1.31
     lake
    1.23
     rocky
    1.17
     mountains
    1.12
     ocean
    1.11
     sea
    1.11
     cave
    1.10
     depths
    1.10
     lakes
    1.10
     lava
    1.09
    Activations Density 0.278%

    No Known Activations