© 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. 26656
    Prev
    Next
    INDEX
    Explanations

    phrases that exhibit a sense of being unrestrained or unbounded

    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
    301
    +0.13
    0.6%
    9
    +0.11
    0.5%
    404
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    404
    +0.13
    0.03
    9
    +0.11
    0.02
    301
    +0.09
    0.02
    Negative Logits
    hops
    -0.86
    groups
    -0.79
    roups
    -0.78
    vertisement
    -0.78
    ovember
    -0.76
    rax
    -0.75
    sters
    -0.74
    akov
    -0.72
    apter
    -0.71
    ington
    -0.70
    POSITIVE LOGITS
     admiration
    1.04
     vigilance
    1.03
     devotion
    1.00
     honesty
    0.99
     optimism
    0.97
     loyalty
    0.94
     reverence
    0.92
     gratitude
    0.91
     sunshine
    0.90
     composure
    0.88
    Activations Density 0.060%

    No Known Activations