© 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. 19317
    Prev
    Next
    INDEX
    Explanations

    mentions of freedom of speech, expression, and related rights and issues

    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
    447
    +0.15
    0.7%
    531
    +0.11
    0.5%
    270
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    225
    +0.15
    0.03
    125
    +0.11
    0.02
    531
    +0.11
    0.02
    Negative Logits
    BALL
    -0.67
     CET
    -0.66
    tom
    -0.64
    VIS
    -0.64
    ARDS
    -0.64
     Gord
    -0.63
    Tele
    -0.63
     Stab
    -0.63
    cit
    -0.62
    git
    -0.61
    POSITIVE LOGITS
     rights
    1.07
     protections
    0.94
     Rights
    0.94
    speech
    0.93
     freedoms
    0.92
     liberties
    0.91
    rights
    0.89
     freedom
    0.83
     advocates
    0.83
     pamph
    0.83
    Activations Density 0.063%

    No Known Activations