© 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. 484
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to geopolitical events and locations

    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
    444
    +0.13
    0.6%
    308
    +0.12
    0.5%
    715
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    444
    +0.13
    0.03
    232
    +0.12
    0.02
    406
    +0.10
    0.02
    Negative Logits
    Wr
    -0.64
     nexus
    -0.63
    LIST
    -0.63
     Actual
    -0.63
    VICE
    -0.62
     juices
    -0.62
     pork
    -0.61
    VO
    -0.58
     compuls
    -0.58
    ++++++++++++++++
    -0.58
    POSITIVE LOGITS
    obile
    1.22
    orrow
    1.17
    otive
    1.16
    puter
    1.15
    edia
    1.12
    obiles
    1.09
    agnetic
    1.08
    icro
    1.07
    atoes
    1.07
    essage
    1.05
    Activations Density 0.065%

    No Known Activations