© 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. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 5459
    Prev
    Next
    INDEX
    Explanations

    phrases related to concerns or worries

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    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
    3
    +0.12
    0.6%
    250
    +0.10
    0.5%
    71
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.12
    0.10
    669
    +0.10
    0.07
    648
    +0.10
    0.06
    Negative Logits
    TL
    -0.85
    fuck
    -0.79
    TM
    -0.75
    \/\/
    -0.73
    Laughs
    -0.72
     FUCK
    -0.71
     shit
    -0.69
    MX
    -0.68
    thing
    -0.67
     Fuck
    -0.66
    POSITIVE LOGITS
     shortcomings
    0.96
     widening
    0.95
     feasibility
    0.93
     potential
    0.92
     disparities
    0.92
     perceptions
    0.90
     proliferation
    0.90
     deficiencies
    0.89
     deepening
    0.87
     inequalities
    0.87
    Activations Density 0.600%

    No Known Activations