© 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 24576
    9. 8-RES_FS24576-JB
    11. 9239
    Prev
    Next
    INDEX
    Explanations

    phrases related to carrying out actions or tasks

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    218
    +0.14
    0.6%
    497
    +0.11
    0.5%
    455
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    218
    +0.14
    0.01
    541
    +0.11
    0.02
    581
    +0.10
    0.01
    Negative Logits
    esa
    -0.82
    lua
    -0.73
     Soup
    -0.66
    oxin
    -0.64
    isma
    -0.64
    enty
    -0.63
    onian
    -0.63
    edia
    -0.62
    username
    -0.60
    ixed
    -0.59
    POSITIVE LOGITS
    uate
    0.84
     tasks
    0.82
     reconnaissance
    0.81
     surveillance
    0.81
     renovations
    0.80
     duties
    0.80
     abortions
    0.78
     executions
    0.77
     repairs
    0.77
     killings
    0.74
    Activations Density 0.030%

    No Known Activations