Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 5164
    Prev
    Next
    INDEX
    Explanations

    phrases related to disseminating information or ideas

    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
    207
    +0.10
    0.5%
    720
    +0.10
    0.5%
    25
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    439
    +0.10
    0.03
    25
    +0.10
    0.03
    92
    +0.09
    0.03
    Negative Logits
    */(
    -0.91
    tarians
    -0.74
    --+
    -0.68
     Zup
    -0.67
    cé
    -0.65
    gerald
    -0.64
     pedest
    -0.62
    girlfriend
    -0.61
    hov
    -0.61
    asus
    -0.60
    POSITIVE LOGITS
    sheets
    1.10
     flyers
    1.06
     leaflets
    1.06
     misinformation
    1.03
     gospel
    1.00
     pamph
    0.96
     disinformation
    0.92
     word
    0.89
     awareness
    0.88
     messages
    0.86
    Activations Density 0.734%

    No Known Activations