© 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. 0-RES-JB
    11. 23132
    Prev
    Next
    INDEX
    Explanations

    mentions of a specific person named Ahmadinejad

    oai_token-act-pair · gpt-3.5-turbo

    mentions of a specific individual named "Jad."

    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
     parachute
    -0.63
     climbing
    -0.62
     carbon
    -0.62
     501
    -0.61
     Personnel
    -0.61
     bigger
    -0.61
     unforeseen
    -0.60
     structural
    -0.59
     Pete
    -0.59
     unexpl
    -0.58
    POSITIVE LOGITS
    jad
    5.25
    llah
    1.66
    jab
    1.29
    hran
    1.26
    hani
    1.23
    mad
    1.16
    jri
    1.14
    jas
    1.07
    JM
    1.06
    jriwal
    1.05
    Activations Density 0.029%

    No Known Activations