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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 17542
    Prev
    Next
    INDEX
    Explanations

    mentions of environmental spills or accidents

    oai_token-act-pair · gpt-3.5-turbo

    references to environmental spills, particularly oil spills

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Gy
    -0.73
     Doctor
    -0.67
    Gy
    -0.66
    Psych
    -0.64
    Art
    -0.63
     Dob
    -0.62
    Doctor
    -0.62
     Classical
    -0.61
     subscribed
    -0.61
    æŃ
    -0.60
    POSITIVE LOGITS
     spill
    3.99
     spills
    2.70
     spilled
    2.40
     spilling
    2.17
     leak
    1.65
     leakage
    1.44
     ripple
    1.41
     overflow
    1.35
     overfl
    1.27
     leaking
    1.24
    Activations Density 0.020%

    No Known Activations