© 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
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 5613
    Prev
    Next
    INDEX
    Explanations

    phrases related to requests or calls to action

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the word "we" and its variations, indicating a focus on collective identity or group discussion

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Watt
    -0.55
    ãĤ¯
    -0.55
     unlaw
    -0.54
     overwhel
    -0.54
     offending
    -0.54
    Reply
    -0.53
     Authority
    -0.53
     injuring
    -0.52
     Azerb
    -0.52
    âĢ¢âĢ¢
    -0.51
    POSITIVE LOGITS
    ibo
    0.98
    bsite
    0.96
    blogs
    0.95
    ird
    0.92
     ourselves
    0.90
    IRD
    0.87
    asel
    0.85
    athered
    0.82
    arily
    0.80
    avers
    0.80
    Activations Density 0.159%

    No Known Activations