© 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. 694
    Prev
    Next
    INDEX
    Explanations

    attributions or causes stated in text

    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
    608
    +0.17
    0.8%
    362
    +0.13
    0.6%
    647
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.17
    0.24
    362
    +0.13
    0.18
    647
    +0.13
    0.13
    Negative Logits
    ura
    -0.92
    anooga
    -0.83
    bats
    -0.83
    lest
    -0.82
    ires
    -0.81
    imensional
    -0.79
    ensable
    -0.76
    okes
    -0.75
    lime
    -0.74
    zon
    -0.74
    POSITIVE LOGITS
     sheer
    1.14
     factors
    0.94
     considerations
    0.92
     pree
    0.89
     inexper
    0.88
     tradition
    0.84
     intuition
    0.82
     geography
    0.81
     advancements
    0.81
     fears
    0.78
    Activations Density 4.945%

    No Known Activations