© 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. 7225
    Prev
    Next
    INDEX
    Explanations

    instances of words related to qualities like "good" and "bad"

    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
    42
    +0.12
    0.6%
    386
    +0.11
    0.5%
    615
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    42
    +0.12
    0.06
    579
    +0.11
    0.04
    628
    +0.10
    0.04
    Negative Logits
    prise
    -0.72
    ulton
    -0.71
     VIDEOS
    -0.65
    iann
    -0.63
    IRE
    -0.63
     Ambrose
    -0.60
     feasibility
    -0.59
     EF
    -0.59
     passage
    -0.59
    risome
    -0.59
    POSITIVE LOGITS
    enough
    0.98
     enough
    0.86
    looking
    0.82
    friends
    0.80
    wired
    0.80
     negotiator
    0.79
     Enough
    0.77
     behaved
    0.76
     undermin
    0.76
     athlet
    0.75
    Activations Density 0.197%

    No Known Activations