© 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. 41588
    Prev
    Next
    INDEX
    Explanations

    words that start with the prefix "pre."

    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
    710
    +0.12
    0.5%
    540
    +0.10
    0.5%
    555
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    710
    +0.12
    0.01
    155
    +0.10
    0.01
    555
    +0.10
    0.01
    Negative Logits
    ãĥīãĥ©ãĤ´ãĥ³
    -0.90
    é¾
    -0.74
    DEV
    -0.74
     GOODMAN
    -0.73
    è¦ļéĨĴ
    -0.72
    loo
    -0.68
    ãĤ¼ãĤ¦ãĤ¹
    -0.68
    onne
    -0.66
    IDS
    -0.64
    BuyableInstoreAndOnline
    -0.63
    POSITIVE LOGITS
    cedented
    1.39
    pared
    1.35
    cision
    1.31
    hens
    1.26
    cise
    1.22
    ferred
    1.17
    vious
    1.15
    viously
    1.11
    cious
    1.05
    empt
    1.03
    Activations Density 0.006%

    No Known Activations