© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 4089
    Prev
    Next
    INDEX
    Explanations

    references to specific characters or individuals labeled as "agents" 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
    527
    +0.11
    0.5%
    218
    +0.11
    0.5%
    427
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    527
    +0.11
    0.02
    280
    +0.11
    0.02
    339
    +0.10
    0.02
    Negative Logits
    ldom
    -0.83
     Parables
    -0.76
     nutrit
    -0.76
    ulz
    -0.73
    uble
    -0.72
    yg
    -0.72
    famous
    -0.72
     "$:/
    -0.72
    ashtra
    -0.72
    jri
    -0.71
    POSITIVE LOGITS
    ij士
    0.79
    agent
    0.75
    SON
    0.69
    IR
    0.69
     Agents
    0.69
    asonic
    0.68
     ATF
    0.68
     EMS
    0.68
    âĸĪâĸĪâĸĪâĸĪ
    0.68
     Briggs
    0.68
    Activations Density 0.582%

    No Known Activations