© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 49152
    9. 8-RES_FS49152-JB
    11. 45501
    Prev
    Next
    INDEX
    Explanations

    the conjunction "and" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    762
    +0.11
    0.5%
    665
    +0.11
    0.5%
    587
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.11
    0.02
    56
    +0.11
    0.01
    718
    +0.10
    0.01
    Negative Logits
    ELD
    -0.70
    ELY
    -0.69
     Dragonbound
    -0.66
    TIME
    -0.65
    YC
    -0.65
    RIC
    -0.65
     Tablet
    -0.63
    Soc
    -0.63
    Joy
    -0.62
     Availability
    -0.61
    POSITIVE LOGITS
     recreate
    1.13
     replicate
    1.10
     convince
    1.04
     emulate
    1.01
     revive
    0.99
     justify
    0.98
     reconcile
    0.97
     negotiate
    0.95
     imitate
    0.95
     persuade
    0.94
    Activations Density 0.020%

    No Known Activations