© 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. 3743
    Prev
    Next
    INDEX
    Explanations

    references to traps and strategies involving deception or manipulation

    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
    517
    +0.11
    0.5%
    169
    +0.10
    0.5%
    581
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    353
    +0.11
    0.01
    613
    +0.10
    0.01
    581
    +0.10
    0.01
    Negative Logits
     reckoning
    -0.67
     pension
    -0.67
    é¾įåĸļ士
    -0.66
     todd
    -0.66
     defaults
    -0.65
     Muse
    -0.64
     unres
    -0.63
     Prix
    -0.61
     default
    -0.61
     Speedway
    -0.59
    POSITIVE LOGITS
    poke
    0.66
     Penguins
    0.66
     Devils
    0.65
    species
    0.64
    asive
    0.64
    inqu
    0.63
    ttes
    0.63
    vous
    0.62
     Experts
    0.62
     Calls
    0.61
    Activations Density 0.264%

    No Known Activations