© 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 12288
    9. 8-RES_FS12288-JB
    11. 9744
    Prev
    Next
    INDEX
    Explanations

    information related to technical issues and news updates, particularly focusing on reports of problems or challenges with online platforms

    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
    481
    +0.30
    1.5%
    447
    +0.26
    1.3%
    470
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.30
    0.14
    481
    +0.26
    0.13
    575
    +0.12
    0.10
    Negative Logits
    )).
    -0.68
    ĪĴ
    -0.68
     guiActive
    -0.65
    ĨĴ
    -0.59
    ?).
    -0.59
    )."
    -0.57
     ?)
    -0.57
    ))
    -0.56
    catentry
    -0.56
    ļé
    -0.56
    POSITIVE LOGITS
    .
    0.75
     because
    0.74
     nowadays
    0.72
     lately
    0.67
     thanks
    0.66
     besides
    0.61
     although
    0.58
     since
    0.58
     dating
    0.58
    ;
    0.57
    Activations Density 1.781%

    No Known Activations