© 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. 4-RES-JB
    11. 10196
    Prev
    Next
    INDEX
    Explanations

    references to the term "Merciful" or similar

    oai_token-act-pair · gpt-3.5-turbo

    references to a character or theme related to "Merc."

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

    No Comments

    Negative Logits
    doors
    -0.83
    FORMATION
    -0.82
    åĤ
    -0.80
    Madison
    -0.72
    Lay
    -0.71
    FER
    -0.71
    eking
    -0.71
    WARE
    -0.70
    VICE
    -0.70
     Ancients
    -0.69
    POSITIVE LOGITS
    iless
    1.28
    enaries
    1.27
    ifully
    1.07
    uria
    1.01
    iful
    0.92
     simultane
    0.89
    opol
    0.88
    urious
    0.87
    adian
    0.87
    ues
    0.85
    Activations Density 0.008%

    No Known Activations