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. 3-RES-JB
    6. 19471
    Prev
    Next
    INDEX
    Explanations

    references to the militant group Hezbollah

    oai_token-act-pair · gpt-3.5-turbo

    references to the militant group Hezbollah

    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
    juries
    -0.87
    mson
    -0.83
     Jenner
    -0.72
     Scot
    -0.72
     Edinburgh
    -0.68
     Nottingham
    -0.66
     Wilde
    -0.65
     Compton
    -0.64
    à©
    -0.64
    raught
    -0.63
    POSITIVE LOGITS
     Hezbollah
    0.92
     operatives
    0.90
     affiliation
    0.87
    ollah
    0.83
     affili
    0.83
     mosqu
    0.78
     dictatorship
    0.76
     holdings
    0.76
     strongh
    0.76
     militias
    0.75
    Activations Density 0.004%

    No Known Activations