© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 10540
    Prev
    Next
    INDEX
    Explanations

    references to legal situations and procedures

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    204
    +0.11
    0.5%
    352
    +0.10
    0.5%
    320
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.11
    0.09
    351
    +0.10
    0.05
    469
    +0.10
    0.05
    Negative Logits
    ggles
    -0.68
    avoid
    -0.66
    rather
    -0.65
    icon
    -0.65
    okingly
    -0.64
    Discussion
    -0.63
    yrinth
    -0.61
     NEVER
    -0.61
    Few
    -0.59
    yssey
    -0.59
    POSITIVE LOGITS
     anymore
    1.64
     nor
    1.33
     whatsoever
    1.06
     anything
    1.04
     any
    1.02
     altogether
    1.00
     anybody
    0.91
     due
    0.91
     properly
    0.88
     anyone
    0.87
    Activations Density 0.746%

    No Known Activations