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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 10-RES_POST_32K-OAI
    11. 30212
    Prev
    Next
    INDEX
    Explanations

    phrases indicating denial or negation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    718
    +0.15
    0.7%
    608
    +0.11
    0.5%
    390
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    616
    +0.15
    0.04
    608
    +0.11
    0.04
    390
    +0.10
    0.03
    Negative Logits
    projects
    -1.08
    devices
    -1.06
    breakers
    -1.01
    units
    -1.01
    agents
    -1.01
    types
    -0.99
    values
    -0.97
    eners
    -0.96
    parts
    -0.95
    results
    -0.94
    POSITIVE LOGITS
     indication
    1.12
     question
    1.08
     denying
    1.03
     possibility
    1.02
     way
    1.00
     reason
    0.99
     shortage
    0.98
     doubt
    0.98
     suggestion
    0.94
     difference
    0.94
    Activations Density 0.103%

    No Known Activations