© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5418
    Prev
    Next
    INDEX
    Explanations

    phrases related to the existence and actions of entities in various contexts

    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
    516
    +0.12
    0.6%
    689
    +0.12
    0.5%
    86
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    139
    +0.12
    0.11
    551
    +0.12
    0.08
    477
    +0.10
    0.06
    Negative Logits
     Crazy
    -0.83
     gonna
    -0.73
     digging
    -0.70
     twisting
    -0.70
     gravy
    -0.70
     goofy
    -0.69
     fuckin
    -0.68
     gotta
    -0.68
     whispering
    -0.67
     mobs
    -0.67
    POSITIVE LOGITS
    atisf
    0.91
     exceeds
    0.90
    tains
    0.86
    byter
    0.86
     subsequently
    0.81
     deems
    0.80
     lawfully
    0.79
     receives
    0.78
    displayText
    0.78
     possesses
    0.76
    Activations Density 0.483%

    No Known Activations