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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 26080
    Prev
    Next
    INDEX
    Explanations

    linguistic elements that refer to specific cultural or geographic identities

    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
    71
    +0.12
    0.5%
    468
    +0.10
    0.5%
    728
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    299
    +0.12
    0.01
    154
    +0.10
    0.01
    284
    +0.10
    0.01
    Negative Logits
    isSpecialOrderable
    -0.79
    Downloadha
    -0.74
     CFR
    -0.72
     Macy
    -0.65
    govtrack
    -0.64
    ources
    -0.64
     NCT
    -0.63
    urchase
    -0.62
     redress
    -0.62
     CU
    -0.60
    POSITIVE LOGITS
    wagen
    0.96
    obiles
    0.81
    Delivery
    0.76
    Ģ
    0.76
    «
    0.75
    ismo
    0.75
    obil
    0.75
    ¡
    0.74
    ¶æ
    0.74
    Ĵ
    0.74
    Activations Density 0.011%

    No Known Activations