© 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 49152
    9. 8-RES_FS49152-JB
    11. 37288
    Prev
    Next
    INDEX
    Explanations

    references to specific opioid substances and their impact on addiction

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    481
    +0.15
    0.7%
    195
    +0.11
    0.5%
    288
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    461
    +0.15
    0.01
    712
    +0.11
    0.01
    231
    +0.11
    0.01
    Negative Logits
    elong
    -1.09
    ffen
    -0.82
    eor
    -0.79
    ership
    -0.78
    edom
    -0.77
    anova
    -0.77
    orkshire
    -0.76
    ateur
    -0.76
    wagen
    -0.75
    */(
    -0.74
    POSITIVE LOGITS
     analges
    1.28
     overdose
    1.12
     fentanyl
    1.11
     opioid
    1.06
     epidemic
    1.03
     addiction
    1.01
     medication
    0.99
     medications
    0.97
     drugs
    0.96
     overdoses
    0.96
    Activations Density 0.006%

    No Known Activations