© 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. 5519
    Prev
    Next
    INDEX
    Explanations

    details about testing and safety concerning drinking water and potential contaminants

    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
    218
    +0.10
    0.5%
    74
    +0.10
    0.5%
    357
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    455
    +0.10
    0.11
    218
    +0.10
    0.05
    408
    +0.10
    0.12
    Negative Logits
    Cooldown
    -0.81
     )]
    -0.72
     narrator
    -0.71
     Wem
    -0.70
    }}}
    -0.70
    ffe
    -0.70
     sermon
    -0.70
     partName
    -0.69
    ruary
    -0.69
     generously
    -0.68
    POSITIVE LOGITS
     contaminants
    1.23
     pathogens
    1.08
     abnormalities
    1.08
     toxins
    1.07
     violations
    1.04
     abnormal
    1.03
     wrongdoing
    1.02
     malware
    1.01
     suspicious
    1.00
     contamination
    0.99
    Activations Density 3.098%

    No Known Activations