© 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
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 1518
    Prev
    Next
    INDEX
    Explanations

    words related to social and political issues, including health policy, economics, and systemic discrimination

    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
    36
    +0.24
    1.2%
    481
    +0.13
    0.6%
    393
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.24
    0.12
    263
    +0.13
    0.06
    159
    +0.11
    0.06
    Negative Logits
    antry
    -0.75
    arsity
    -0.74
    Arcade
    -0.73
     honors
    -0.67
    <?
    -0.66
    grounds
    -0.66
    Angel
    -0.65
     OWN
    -0.65
    Own
    -0.64
    Instruct
    -0.62
    POSITIVE LOGITS
     lot
    1.18
    cknowled
    1.07
     handful
    1.06
     few
    1.03
     majority
    1.02
     combination
    0.99
    usterity
    0.98
     glance
    0.96
     lack
    0.94
     plethora
    0.93
    Activations Density 0.195%

    No Known Activations