Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 231
    Prev
    Next
    INDEX
    Explanations

    phrases related to personal beliefs or actions that have significant consequences

    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
    447
    +0.14
    0.6%
    2
    +0.12
    0.5%
    349
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    604
    +0.14
    0.06
    566
    +0.12
    0.07
    750
    +0.10
    0.06
    Negative Logits
    currently
    -0.78
    enei
    -0.74
    tis
    -0.73
    arel
    -0.70
    mania
    -0.69
    linked
    -0.67
    held
    -0.64
    weather
    -0.63
     currently
    -0.63
     presently
    -0.62
    POSITIVE LOGITS
     yesterday
    0.81
     beforehand
    0.79
     LAST
    0.78
     Doct
    0.71
     last
    0.71
     wrong
    0.68
     Pike
    0.68
    nesday
    0.67
     me
    0.67
     alot
    0.67
    Activations Density 0.707%

    No Known Activations