Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5155
    Prev
    Next
    INDEX
    Explanations

    percentages in data

    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
    147
    +0.11
    0.5%
    285
    +0.09
    0.4%
    466
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    742
    +0.11
    0.05
    147
    +0.09
    0.06
    743
    +0.09
    0.05
    Negative Logits
    ãĤ©
    -0.64
    beit
    -0.62
     compan
    -0.60
     shaped
    -0.57
     nurs
    -0.56
    worldly
    -0.56
     detrim
    -0.55
     evid
    -0.55
     cradle
    -0.55
    ham
    -0.54
    POSITIVE LOGITS
    +.
    0.88
    +)
    0.80
     ABV
    0.79
    +,
    0.75
    -+
    0.71
    xual
    0.70
    ablishment
    0.69
    payer
    0.67
     chance
    0.66
    azo
    0.66
    Activations Density 0.191%

    No Known Activations