© 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. 4984
    Prev
    Next
    INDEX
    Explanations

    references to ordnance or military equipment

    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
    31
    +0.10
    0.4%
    383
    +0.10
    0.4%
    275
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    31
    +0.10
    0.02
    7
    +0.10
    0.01
    383
    +0.10
    0.01
    Negative Logits
    pmwiki
    -0.62
     Bind
    -0.62
     SEN
    -0.62
    vous
    -0.60
    burn
    -0.59
     Dull
    -0.59
    Ö¼
    -0.58
     Rust
    -0.57
    ted
    -0.57
    lasses
    -0.57
    POSITIVE LOGITS
    nance
    1.35
    inary
    1.11
    inances
    1.02
    inarily
    0.98
    inance
    0.94
    eal
    0.82
    onom
    0.79
    istry
    0.76
    oser
    0.74
    ered
    0.74
    Activations Density 0.083%

    No Known Activations