© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 1847
    Prev
    Next
    INDEX
    Explanations

    words related to honor, particularly within a negative or violent context

    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
    590
    +0.12
    0.5%
    288
    +0.11
    0.5%
    7
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    590
    +0.12
    0.02
    424
    +0.11
    0.01
    165
    +0.11
    0.01
    Negative Logits
    arre
    -0.86
    hesive
    -0.72
     Ange
    -0.69
    _>
    -0.68
    roads
    -0.67
     Lans
    -0.66
     Reloaded
    -0.64
    ept
    -0.63
     tyres
    -0.60
    FH
    -0.60
    POSITIVE LOGITS
    ably
    1.22
    ific
    1.18
     honors
    0.97
     honored
    0.95
     honoured
    0.93
    abilia
    0.91
     honor
    0.84
     bestowed
    0.83
    orable
    0.82
     honorable
    0.81
    Activations Density 0.041%

    No Known Activations