© 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. Attention Out - 32k
    5. 3-ATT_32K-OAI
    6. 4023
    Prev
    Next
    INDEX
    Explanations

    references to political parties and affiliations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-attn-out-SAEs/v5_32k_layer_3
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.3.hook_attn_out
    Hook Layer
    3
    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
    532
    +0.11
    0.5%
    713
    +0.10
    0.5%
    602
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    476
    +0.11
    0.04
    596
    +0.10
    0.05
    749
    +0.10
    0.05
    Negative Logits
     Reborn
    -0.67
    pring
    -0.66
    allo
    -0.65
    issan
    -0.64
    £ı
    -0.64
    iga
    -0.63
     ACTIONS
    -0.62
    rative
    -0.61
     Fired
    -0.60
    bard
    -0.59
    POSITIVE LOGITS
    sol
    0.79
    sing
    0.70
    Sol
    0.69
     IPM
    0.64
    eur
    0.64
     horizont
    0.63
    ');
    0.62
     Sche
    0.60
    plet
    0.59
     unbeliev
    0.57
    Activations Density 0.655%

    No Known Activations