© 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. 19316
    Prev
    Next
    INDEX
    Explanations

    contractions of "is" with another word following it

    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
    608
    +0.16
    0.7%
    241
    +0.10
    0.5%
    126
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    126
    +0.16
    0.03
    398
    +0.10
    0.03
    608
    +0.10
    0.03
    Negative Logits
    roit
    -0.78
    eal
    -0.76
    ares
    -0.71
    soever
    -0.64
    icut
    -0.61
     wastes
    -0.61
    agos
    -0.60
     approves
    -0.59
     implements
    -0.59
    ally
    -0.58
    POSITIVE LOGITS
     gotta
    1.08
     been
    1.04
     plenty
    1.02
     gonna
    0.96
     always
    0.93
     no
    0.84
     nothing
    0.83
     definitely
    0.81
     still
    0.80
     lots
    0.80
    Activations Density 0.048%

    No Known Activations