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. 8-RES-JB
    6. 17938
    Prev
    Next
    INDEX
    Explanations

    instances where communication happens via a specific platform or method, such as "via Facebook" or "via mail."

    oai_token-act-pair · gpt-3.5-turbo

    references to methods of communication or transmission

    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
    stal
    -0.85
    >>\
    -0.76
    stad
    -0.75
    lah
    -0.73
    thal
    -0.69
    stall
    -0.68
    inho
    -0.68
    aed
    -0.67
    ETF
    -0.67
    âĢķ
    -0.66
    POSITIVE LOGITS
     Skype
    0.91
     email
    0.88
     Bluetooth
    0.79
     telephone
    0.79
     intermedi
    0.75
     Via
    0.71
     HDMI
    0.71
     phone
    0.70
     SSH
    0.70
    pass
    0.70
    Activations Density 0.055%

    No Known Activations