© 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. 24006
    Prev
    Next
    INDEX
    Explanations

    terms related to symmetry and asymmetry

    oai_token-act-pair · gpt-4o-miniTriggered 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
    715
    +0.13
    0.6%
    481
    +0.13
    0.6%
    503
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    503
    +0.13
    0.01
    288
    +0.13
    0.02
    539
    +0.12
    0.01
    Negative Logits
     ado
    -0.81
     guesses
    -0.76
    house
    -0.67
    gling
    -0.65
    OWS
    -0.63
     liner
    -0.62
     Undead
    -0.61
     Juven
    -0.60
     MIA
    -0.59
    Footnote
    -0.59
    POSITIVE LOGITS
    etrical
    1.91
    etric
    1.84
    etry
    1.49
    etr
    1.41
    eties
    1.11
    ethy
    1.09
    olitics
    1.05
    et
    1.02
    ymm
    1.01
    ility
    1.00
    Activations Density 0.032%

    No Known Activations