© 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. 9787
    Prev
    Next
    INDEX
    Explanations

    news events or incidents, such as crimes, accidents, and police reports

    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
    +0.55
    2.9%
    266
    +0.13
    0.7%
    516
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.55
    0.02
    266
    +0.13
    0.12
    516
    +0.10
    0.07
    Negative Logits
     ç¥ŀ
    -0.79
     optim
    -0.78
    optim
    -0.74
     UNESCO
    -0.71
    çļ
    -0.66
     emulation
    -0.59
     releg
    -0.59
    女
    -0.58
     dividends
    -0.58
     trillions
    -0.56
    POSITIVE LOGITS
     burglary
    1.09
     police
    1.08
    Police
    1.03
     robbery
    1.01
    police
    1.01
     burgl
    1.00
     altercation
    0.98
     deputies
    0.98
     gunfire
    0.98
     detectives
    0.97
    Activations Density 1.887%

    No Known Activations