© 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 24576
    9. 8-RES_FS24576-JB
    11. 1957
    Prev
    Next
    INDEX
    Explanations

    evidence of suspicious actions or events, such as bruises, spoilers, errors, drugs, locks, alarms, signs, tanks, and windows

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    447
    +0.21
    1.0%
    282
    +0.11
    0.5%
    575
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.21
    0.08
    258
    +0.11
    0.08
    114
    +0.10
    0.06
    Negative Logits
    LESS
    -0.65
    Leader
    -0.63
     Travels
    -0.62
     Orth
    -0.62
    Studies
    -0.62
    Station
    -0.60
    times
    -0.58
    Rate
    -0.58
    ãĤ¨ãĥ«
    -0.56
    ther
    -0.55
    POSITIVE LOGITS
    creen
    1.22
    paces
    1.17
    hips
    1.13
    peed
    1.11
    cale
    1.10
    pace
    1.09
    heet
    1.07
    mith
    1.07
    poons
    1.06
    cape
    1.04
    Activations Density 0.420%

    No Known Activations