© 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. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 5259
    Prev
    Next
    INDEX
    Explanations

    references to identities and names in the context of victims and suspects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    121
    +0.14
    0.6%
    697
    +0.10
    0.4%
    590
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.14
    0.10
    121
    +0.10
    0.08
    264
    +0.10
    0.03
    Negative Logits
    liga
    -0.72
     Enhancement
    -0.71
    rite
    -0.71
    lance
    -0.70
     Efficiency
    -0.70
    wcsstore
    -0.69
    TYPE
    -0.67
    dayName
    -0.67
     specialization
    -0.66
    entious
    -0.66
    POSITIVE LOGITS
     deceased
    1.31
     victims
    1.24
     slain
    1.19
     accused
    1.10
     murdered
    1.10
     detained
    1.08
     alleged
    1.08
     defendants
    1.07
     convicted
    1.06
     suspects
    1.05
    Activations Density 1.518%

    No Known Activations