© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3753
    Prev
    Next
    INDEX
    Explanations

    personal beliefs or perspectives in textual content

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    581
    +0.12
    0.6%
    53
    +0.12
    0.5%
    674
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    53
    +0.12
    0.03
    54
    +0.12
    0.03
    581
    +0.09
    0.03
    Negative Logits
    lyak
    -0.82
    gart
    -0.72
    odan
    -0.65
    Fif
    -0.63
    RAFT
    -0.62
    sted
    -0.62
    FER
    -0.61
    uca
    -0.59
    YN
    -0.58
    Esc
    -0.58
    POSITIVE LOGITS
    pires
    0.98
     imaginable
    0.92
     except
    0.82
    abilia
    0.82
    except
    0.80
     entails
    0.74
    ãĤ¨ãĥ«
    0.72
     happ
    0.72
    necessary
    0.71
     transpired
    0.70
    Activations Density 0.160%

    No Known Activations