© 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. 3-RES-JB
    11. 14778
    Prev
    Next
    INDEX
    Explanations

    references to the social media platform Tumblr

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the social media platform Tumblr

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    eering
    -0.74
    undai
    -0.68
     shotguns
    -0.66
    pse
    -0.66
    kj
    -0.62
    utherford
    -0.62
    bid
    -0.60
    imentary
    -0.59
     Adin
    -0.59
     treaty
    -0.59
    POSITIVE LOGITS
    umblr
    1.00
     Tumblr
    0.93
    Reddit
    0.80
    Tumblr
    0.79
     Doodle
    0.79
    olor
    0.79
     Reader
    0.77
    verse
    0.77
     Username
    0.75
     Flickr
    0.74
    Activations Density 0.005%

    No Known Activations