© 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. 9511
    Prev
    Next
    INDEX
    Explanations

    information related to movies, films, and productions

    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
    373
    +0.22
    1.1%
    198
    +0.08
    0.4%
    655
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    344
    +0.22
    0.06
    566
    +0.08
    0.04
    198
    +0.08
    0.03
    Negative Logits
    antis
    -0.70
     Thing
    -0.68
    abuse
    -0.66
    truth
    -0.65
     trolls
    -0.64
    got
    -0.60
     tricks
    -0.59
    hemy
    -0.59
    ickr
    -0.58
     crappy
    -0.57
    POSITIVE LOGITS
     approximately
    0.78
     estimated
    0.70
    Tickets
    0.69
     culmin
    0.66
     totaled
    0.66
     headquartered
    0.65
     comprised
    0.65
    unda
    0.64
     scheduled
    0.63
     overseen
    0.63
    Activations Density 0.627%

    No Known Activations