© 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. 9762
    Prev
    Next
    INDEX
    Explanations

    the word "ready"

    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
    388
    +0.10
    0.5%
    240
    +0.10
    0.5%
    213
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    240
    +0.10
    0.03
    222
    +0.10
    0.02
    668
    +0.09
    0.02
    Negative Logits
    uo
    -0.76
    loo
    -0.69
    oret
    -0.68
    regate
    -0.67
    Publisher
    -0.66
    adish
    -0.65
    VERTISEMENT
    -0.61
     sting
    -0.60
    eatures
    -0.60
    orie
    -0.60
    POSITIVE LOGITS
    Ready
    0.90
     ready
    0.83
    housing
    0.81
     Ready
    0.79
    itude
    0.75
    arity
    0.72
    wired
    0.71
    made
    0.70
    ÃįÃį
    0.70
     ripe
    0.69
    Activations Density 0.058%

    No Known Activations