© 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. 4772
    Prev
    Next
    INDEX
    Explanations

    structured elements related to certifications and exams

    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
    165
    +0.11
    0.5%
    679
    +0.10
    0.5%
    711
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    679
    +0.11
    0.23
    739
    +0.10
    0.05
    640
    +0.09
    0.21
    Negative Logits
     figur
    -0.83
     scenes
    -0.66
     senseless
    -0.65
     prolet
    -0.64
     landscapes
    -0.64
     rooft
    -0.63
     bloodshed
    -0.61
     hygiene
    -0.60
     toile
    -0.59
    atomic
    -0.59
    POSITIVE LOGITS
     triggered
    0.75
    ortunate
    0.73
     activated
    0.72
     whereby
    0.71
    enhagen
    0.69
     wherein
    0.68
    pkg
    0.67
    tesy
    0.67
    amacare
    0.66
    ogg
    0.65
    Activations Density 7.746%

    No Known Activations