© 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. 9522
    Prev
    Next
    INDEX
    Explanations

    proper nouns and terms related to government, legal matters, and industry

    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
    254
    +0.12
    0.6%
    481
    +0.12
    0.5%
    151
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.12
    0.07
    224
    +0.12
    0.05
    254
    +0.10
    0.05
    Negative Logits
    %%%%
    -0.72
    wed
    -0.67
    quickShipAvailable
    -0.64
    alam
    -0.63
    lander
    -0.61
    wrapper
    -0.60
    vet
    -0.59
    phas
    -0.59
    pmwiki
    -0.58
    ivalry
    -0.58
    POSITIVE LOGITS
     via
    0.99
     coffers
    0.94
     detriment
    0.74
     tune
    0.72
     doorstep
    0.70
     needy
    0.70
    rolet
    0.69
     recipients
    0.69
    umers
    0.69
     ASAP
    0.66
    Activations Density 0.475%

    No Known Activations