© 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. 5261
    Prev
    Next
    INDEX
    Explanations

    the word "general" and its variations

    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
    373
    +0.19
    0.9%
    317
    +0.12
    0.5%
    13
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    622
    +0.19
    0.02
    613
    +0.12
    0.02
    737
    +0.10
    0.02
    Negative Logits
    arcity
    -0.67
     Feld
    -0.66
     faked
    -0.66
    zl
    -0.63
     inflamm
    -0.63
     Glacier
    -0.61
     Stainless
    -0.60
     underestimated
    -0.58
     Helic
    -0.58
     exagger
    -0.57
    POSITIVE LOGITS
    ciples
    0.71
    displayText
    0.71
    aga
    0.70
    enture
    0.69
    aez
    0.68
    etta
    0.65
    notation
    0.64
    fter
    0.64
    Rules
    0.63
    iola
    0.63
    Activations Density 0.129%

    No Known Activations