© 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 1536
    9. 8-RES_FS1536-JB
    11. 268
    Prev
    Next
    INDEX
    Explanations

    numbers or comparisons

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    539
    +0.10
    0.5%
    662
    +0.10
    0.4%
    178
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    539
    +0.10
    0.08
    296
    +0.10
    0.10
    647
    +0.10
    0.07
    Negative Logits
     Became
    -0.61
    Limited
    -0.60
    ointed
    -0.59
    fw
    -0.58
    whe
    -0.58
    bit
    -0.58
    iverse
    -0.55
    awar
    -0.55
    Premium
    -0.55
    iol
    -0.54
    POSITIVE LOGITS
     usual
    1.17
     previous
    1.15
     predecessors
    1.10
     traditional
    1.06
     typical
    1.04
     counterparts
    1.03
     conventional
    0.99
    usual
    0.99
     others
    0.95
     predecessor
    0.89
    Activations Density 2.236%

    No Known Activations