© 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. 6-RES-JB
    11. 20413
    Prev
    Next
    INDEX
    Explanations

    keywords related to communication tools and functions such as Skype, chat, contacts, video calls, calendars, messaging, and notes

    oai_token-act-pair · gpt-3.5-turbo

    features related to communication and organization tools

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 6-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    abama
    -0.80
    judicial
    -0.77
    stood
    -0.72
    eln
    -0.70
    iltration
    -0.69
    jandro
    -0.69
    osph
    -0.68
    oil
    -0.68
    fitting
    -0.67
    ilitary
    -0.67
    POSITIVE LOGITS
     Gmail
    0.90
     playlist
    0.90
     Dropbox
    0.87
     folders
    0.82
     Analytics
    0.82
     subreddits
    0.81
     apps
    0.81
     Cortana
    0.80
     Settings
    0.79
     URLs
    0.79
    Activations Density 0.377%

    No Known Activations