© 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. 45842
    Prev
    Next
    INDEX
    Explanations

    references to the word "Re" or variations thereof

    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
    439
    +0.12
    0.5%
    706
    +0.11
    0.5%
    710
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    439
    +0.12
    0.03
    710
    +0.11
    0.02
    706
    +0.10
    0.02
    Negative Logits
    NESS
    -0.72
     Pebble
    -0.68
    DragonMagazine
    -0.64
     handwriting
    -0.63
     Northern
    -0.59
     Icelandic
    -0.59
     pall
    -0.59
     CPR
    -0.58
     trem
    -0.58
     Korra
    -0.58
    POSITIVE LOGITS
    levant
    1.33
    usable
    1.24
    uben
    1.24
    asonable
    1.21
    ceive
    1.17
    gex
    1.15
    claim
    1.14
    versible
    1.14
    acher
    1.13
    actor
    1.10
    Activations Density 0.014%

    No Known Activations