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. 0-RES-JB
    6. 21520
    Prev
    Next
    INDEX
    Explanations

    mentions of different types of personal records such as rental records and credit card records, as well as locations like apartments

    oai_token-act-pair · gpt-3.5-turbo

    references to apartments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ils
    -0.66
    Stew
    -0.64
    Force
    -0.63
    atics
    -0.63
    GA
    -0.62
    bull
    -0.60
    worm
    -0.60
    WIN
    -0.59
     scientifically
    -0.59
    Graham
    -0.59
    POSITIVE LOGITS
     apartment
    3.44
     apartments
    2.65
     condo
    2.17
    artment
    1.65
     condos
    1.64
     bedroom
    1.53
     roomm
    1.51
    ominium
    1.50
     dorm
    1.49
     roommate
    1.48
    Activations Density 0.024%

    No Known Activations