© 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. 750
    Prev
    Next
    INDEX
    Explanations

    names related to individuals in a political context, particularly involving North Korea

    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
    9
    +0.10
    0.5%
    443
    +0.10
    0.5%
    420
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    639
    +0.10
    0.03
    9
    +0.10
    0.02
    443
    +0.10
    0.02
    Negative Logits
     replay
    -0.65
     MOT
    -0.64
     multipl
    -0.64
     destro
    -0.64
     Digest
    -0.63
     macros
    -0.62
     REAL
    -0.61
     ditch
    -0.61
     defin
    -0.61
    ATTLE
    -0.61
    POSITIVE LOGITS
    sung
    1.16
    jin
    1.09
    Yang
    1.04
    wei
    1.04
    Hong
    1.03
    kai
    0.99
    Yu
    0.97
    Yan
    0.97
    hun
    0.97
    Tai
    0.96
    Activations Density 0.040%

    No Known Activations