© 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. 9742
    Prev
    Next
    INDEX
    Explanations

    terms related to legal matters and criminal activities

    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
    640
    +0.27
    1.3%
    447
    +0.20
    1.0%
    160
    +0.19
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.27
    0.14
    160
    +0.20
    0.14
    107
    +0.19
    0.12
    Negative Logits
     describ
    -0.84
     neighb
    -0.80
     spitting
    -0.76
     sculpt
    -0.75
     frontline
    -0.71
     racing
    -0.70
     quir
    -0.70
     earthqu
    -0.68
     lucky
    -0.68
     nightly
    -0.68
    POSITIVE LOGITS
    Advertisements
    1.38
    References
    1.30
    Advertisement
    1.29
    RAW
    1.28
    CONCLUS
    1.27
    Discussion
    1.27
    âĨij
    1.26
    ©
    1.24
    Newsletter
    1.23
    Trivia
    1.23
    Activations Density 0.469%

    No Known Activations