© 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. 884
    Prev
    Next
    INDEX
    Explanations

    phrases related to ethics and economic policies, especially with a focus on self-interest and financial implications

    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
    480
    +0.51
    2.6%
    266
    +0.09
    0.5%
    140
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.51
    0.00
    266
    +0.09
    0.11
    351
    +0.08
    0.04
    Negative Logits
     Boll
    -0.57
    mittedly
    -0.50
    PRE
    -0.49
    COMPLE
    -0.49
    Firstly
    -0.48
    QB
    -0.48
    oiler
    -0.48
    Spoiler
    -0.47
    eve
    -0.47
    gression
    -0.47
    POSITIVE LOGITS
     etc
    0.71
    cellaneous
    0.68
    azeera
    0.61
     assorted
    0.59
     attachments
    0.53
    ))))
    0.53
    )).
    0.52
     friendships
    0.52
     tattoos
    0.52
     punishments
    0.51
    Activations Density 1.977%

    No Known Activations