© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 4749
    Prev
    Next
    INDEX
    Explanations

    various forms and contexts of the word "attack."

    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
    288
    +0.12
    0.6%
    448
    +0.11
    0.5%
    531
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    531
    +0.12
    0.05
    288
    +0.11
    0.05
    601
    +0.11
    0.02
    Negative Logits
     Upton
    -0.70
    aple
    -0.64
     Transfer
    -0.62
     Zot
    -0.61
     interchange
    -0.60
    udeau
    -0.59
    åħī
    -0.58
     tidy
    -0.55
     Cups
    -0.55
     veins
    -0.55
    POSITIVE LOGITS
     against
    1.08
     waged
    0.94
    against
    0.93
     perpetrated
    0.89
     targeting
    0.85
    ategory
    0.85
     upon
    0.81
    ivated
    0.80
     aimed
    0.80
    intosh
    0.78
    Activations Density 0.258%

    No Known Activations