© 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 3072
    9. 8-RES_FS3072-JB
    11. 2958
    Prev
    Next
    INDEX
    Explanations

    words related to persistence or continuity

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    458
    +0.10
    0.5%
    599
    +0.09
    0.4%
    459
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    458
    +0.10
    0.07
    38
    +0.09
    0.06
    599
    +0.09
    0.05
    Negative Logits
    insula
    -0.72
     Stim
    -0.67
     elimination
    -0.66
    ASE
    -0.63
    alties
    -0.62
    stice
    -0.61
    ongyang
    -0.60
    iHUD
    -0.60
    ridges
    -0.60
    agus
    -0.59
    POSITIVE LOGITS
    birth
    1.17
     intact
    0.96
    born
    0.91
     unanswered
    0.86
     reeling
    0.79
     undecided
    0.78
     alive
    0.77
     extant
    0.76
     lifes
    0.73
     technically
    0.71
    Activations Density 0.460%

    No Known Activations