© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 43129
    Prev
    Next
    INDEX
    Explanations

    demonstrative pronouns and their usage in context

    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
    709
    +0.12
    0.5%
    560
    +0.11
    0.5%
    159
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    159
    +0.12
    0.03
    560
    +0.11
    0.02
    662
    +0.11
    0.03
    Negative Logits
    icons
    -0.84
    ahs
    -0.75
    emale
    -0.73
    cycles
    -0.73
    english
    -0.69
    ocene
    -0.69
    amia
    -0.66
    hens
    -0.65
    osponsors
    -0.65
    ãĤ¨ãĥ«
    -0.62
    POSITIVE LOGITS
     considerations
    0.84
     complication
    0.79
     reasoning
    0.75
    ,
    0.74
     complicate
    0.72
     uncertainty
    0.72
     fact
    0.71
     scenario
    0.70
     discrepancy
    0.70
     limitation
    0.69
    Activations Density 0.049%

    No Known Activations