© 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 12288
    9. 8-RES_FS12288-JB
    11. 403
    Prev
    Next
    INDEX
    Explanations

    phrases related to speaking out against various issues, including discrimination, violence, and injustice

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    447
    +0.26
    1.2%
    481
    +0.20
    1.0%
    445
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    247
    +0.26
    0.04
    729
    +0.20
    0.04
    666
    +0.10
    0.04
    Negative Logits
    ipment
    -0.67
    plet
    -0.65
    ixture
    -0.60
    ffee
    -0.59
    ãĥ¼ãĥĨãĤ£
    -0.58
     fly
    -0.57
     orbits
    -0.57
    amation
    -0.57
     Ham
    -0.57
     Landing
    -0.57
    POSITIVE LOGITS
     against
    1.01
     Against
    0.92
    against
    0.86
     responsibly
    0.84
     forcefully
    0.82
     financially
    0.82
     peacefully
    0.81
     anonymously
    0.80
     boldly
    0.80
     decisively
    0.79
    Activations Density 0.437%

    No Known Activations