© 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. 9375
    Prev
    Next
    INDEX
    Explanations

    references to physical openings or gaps, specifically focusing on holes

    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
    35
    +0.22
    1.0%
    151
    +0.11
    0.5%
    515
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    35
    +0.22
    0.03
    211
    +0.11
    0.01
    40
    +0.11
    0.02
    Negative Logits
    cia
    -0.83
    ullivan
    -0.71
    NetMessage
    -0.69
     livest
    -0.67
    vez
    -0.67
    Bal
    -0.66
    against
    -0.65
    uscript
    -0.64
    voice
    -0.64
    Revolution
    -0.64
    POSITIVE LOGITS
     hole
    1.26
     holes
    1.23
     drilled
    1.16
    hole
    1.16
     Hole
    1.12
     dug
    1.04
    holes
    0.99
     punched
    0.90
    pots
    0.89
     poked
    0.85
    Activations Density 0.023%

    No Known Activations