© 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 768
    9. 8-RES_FS768-JB
    11. 647
    Prev
    Next
    INDEX
    Explanations

    names and terms related to politics, sports, and events

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    373
    +0.12
    0.6%
    184
    +0.12
    0.6%
    638
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.12
    -0.09
    284
    +0.12
    0.15
    232
    +0.10
    0.15
    Negative Logits
     chains
    -0.69
     ILCS
    -0.60
    MRI
    -0.58
    gerald
    -0.56
     unwelcome
    -0.54
     uneasy
    -0.54
     hurd
    -0.53
     chain
    -0.53
    REE
    -0.52
    ³³³³³³³³
    -0.52
    POSITIVE LOGITS
    zsche
    0.79
    ilus
    0.73
    onyms
    0.68
    vana
    0.66
    anski
    0.66
    achine
    0.65
    ukong
    0.63
    ãĤ»
    0.63
    anyahu
    0.62
    manship
    0.60
    Activations Density 5.047%

    No Known Activations