Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 46182
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "the"

    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
    362
    +0.16
    0.7%
    393
    +0.15
    0.7%
    36
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.16
    0.06
    393
    +0.15
    0.04
    636
    +0.10
    0.03
    Negative Logits
    ngth
    -0.70
    Layer
    -0.67
    AppData
    -0.65
    aurus
    -0.65
    cone
    -0.64
    LET
    -0.63
    Iterator
    -0.63
    lessly
    -0.63
    let
    -0.62
    ãĥīãĥ©
    -0.61
    POSITIVE LOGITS
     topic
    1.22
     eve
    1.19
     basis
    1.18
     behalf
    1.14
     sidelines
    1.12
     occasion
    1.12
     heels
    1.08
     grounds
    1.05
     outskirts
    1.02
     merits
    1.01
    Activations Density 0.104%

    No Known Activations