© 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. 680
    Prev
    Next
    INDEX
    Explanations

    references to online articles from The Huffington Post

    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
    709
    +0.12
    0.6%
    710
    +0.10
    0.5%
    374
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.12
    0.03
    374
    +0.10
    0.01
    288
    +0.10
    0.03
    Negative Logits
    CHR
    -0.76
    clear
    -0.64
     blinded
    -0.63
     Defenders
    -0.60
    ItemImage
    -0.59
     betrayal
    -0.57
    ³³³³
    -0.57
    strate
    -0.56
     Cage
    -0.56
    relations
    -0.56
    POSITIVE LOGITS
    ington
    1.10
    ield
    0.97
    ixel
    0.94
    alo
    0.94
    olk
    0.94
    IELD
    0.91
    eting
    0.88
    erey
    0.87
    enger
    0.87
    etts
    0.83
    Activations Density 0.106%

    No Known Activations