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. 14674
    Prev
    Next
    INDEX
    Explanations

    phrases indicating legal actions or criminal activities leading to arrest

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "on" in various contexts

    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
    RL
    -0.68
    Except
    -0.62
    acters
    -0.62
     ens
    -0.59
    Applications
    -0.59
    Quantity
    -0.58
    rament
    -0.58
    anan
    -0.58
    hammer
    -0.58
    Publisher
    -0.58
    POSITIVE LOGITS
     behalf
    1.46
    etime
    1.17
     occasion
    1.12
    coming
    1.11
    shore
    1.06
    yx
    1.01
     suspicion
    1.00
    site
    0.93
     weekends
    0.92
     occasions
    0.92
    Activations Density 0.236%

    No Known Activations