© 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. 5406
    Prev
    Next
    INDEX
    Explanations

    phrases related to wastefulness and inefficiency

    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
    460
    +0.10
    0.5%
    61
    +0.09
    0.4%
    727
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    570
    +0.10
    0.01
    351
    +0.09
    0.03
    750
    +0.09
    0.02
    Negative Logits
    tain
    -0.70
    Ń·
    -0.68
    ©¶æ
    -0.68
    onis
    -0.66
    tains
    -0.62
    skirts
    -0.61
    ħĭ
    -0.60
    iggurat
    -0.59
     prevail
    -0.59
     tint
    -0.59
    POSITIVE LOGITS
     opportunities
    0.86
     chances
    0.82
     resources
    0.78
     opportunity
    0.77
    ega
    0.73
     valuable
    0.72
    unction
    0.71
     pity
    0.70
     precious
    0.67
    avorite
    0.67
    Activations Density 0.329%

    No Known Activations