© 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. Attention Out - 32k
    9. 3-ATT_32K-OAI
    11. 4059
    Prev
    Next
    INDEX
    Explanations

    terms related to weakness or inadequacy in various contexts

    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
    497
    +0.17
    0.8%
    420
    +0.13
    0.6%
    745
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    338
    +0.17
    0.03
    401
    +0.13
    0.03
    535
    +0.11
    0.02
    Negative Logits
    IBLE
    -0.79
    çīĪ
    -0.74
    aire
    -0.72
    atari
    -0.71
    OTO
    -0.71
    DCS
    -0.69
    Lago
    -0.65
    oir
    -0.65
    ais
    -0.65
    rail
    -0.64
    POSITIVE LOGITS
    inventoryQuantity
    0.97
    luster
    0.91
     feeble
    0.77
     weak
    0.74
     inhib
    0.72
     comprom
    0.71
     cripp
    0.70
     weaknesses
    0.69
     newsp
    0.69
     due
    0.68
    Activations Density 0.406%

    No Known Activations