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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 10147
    Prev
    Next
    INDEX
    Explanations

    names or terms with similar pronunciation

    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
    603
    +0.12
    0.6%
    473
    +0.11
    0.5%
    301
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    603
    +0.12
    0.04
    535
    +0.11
    0.03
    609
    +0.10
    0.04
    Negative Logits
    ONSORED
    -0.81
    taboola
    -0.71
    ndra
    -0.70
    adr
    -0.68
    Premium
    -0.67
    Spons
    -0.66
    drm
    -0.66
    WI
    -0.65
    Progress
    -0.65
    Developer
    -0.65
    POSITIVE LOGITS
     pronunciation
    1.19
    ciation
    1.18
     spelling
    1.17
     suffix
    1.12
     diction
    1.08
     prefix
    1.07
     phon
    1.03
     meanings
    1.02
     phrase
    1.01
     notation
    1.01
    Activations Density 0.229%

    No Known Activations