© 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. 1-RES-JB
    11. 9497
    Prev
    Next
    INDEX
    Explanations

    social media platform names

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the Pinterest platform and related social media content

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Antar
    -0.65
    åĤ
    -0.65
     former
    -0.65
     Poe
    -0.64
     mild
    -0.63
    clus
    -0.63
     Boll
    -0.63
     Cay
    -0.62
    oning
    -0.62
     Breach
    -0.61
    POSITIVE LOGITS
    Pinterest
    1.45
    acebook
    0.90
    LinkedIn
    0.88
    Reddit
    0.88
    utterstock
    0.85
    cffff
    0.84
    avascript
    0.83
    Tumblr
    0.81
    umblr
    0.79
    vine
    0.79
    Activations Density 0.013%

    No Known Activations