© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 656
    Prev
    Next
    INDEX
    Explanations

    political and societal impacts related to various issues and events

    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
    447
    +0.13
    0.6%
    570
    +0.11
    0.5%
    68
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    570
    +0.13
    0.02
    271
    +0.11
    0.03
    296
    +0.09
    0.03
    Negative Logits
    zar
    -0.75
    çĦ
    -0.73
    atom
    -0.68
    Sham
    -0.65
    BuyableInstoreAndOnline
    -0.64
    zen
    -0.63
    DragonMagazine
    -0.63
    ç«
    -0.63
    OVA
    -0.63
    Plot
    -0.63
    POSITIVE LOGITS
     interstate
    0.93
     havoc
    0.93
     adversely
    0.93
     livelihood
    0.86
    uate
    0.85
     peoples
    0.84
     commerce
    0.83
     perceptions
    0.80
     morale
    0.79
     negatively
    0.78
    Activations Density 0.118%

    No Known Activations