© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 9330
    Prev
    Next
    INDEX
    Explanations

    people or entities being blamed by someone in various contexts for different situations or events

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "blamed" in various contexts

    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
    marked
    -0.89
    ymph
    -0.79
    mark
    -0.77
    cre
    -0.75
    edom
    -0.75
    orld
    -0.74
    mental
    -0.74
    visors
    -0.73
    intend
    -0.73
    arent
    -0.72
    POSITIVE LOGITS
     blames
    0.77
     blamed
    0.77
     blaming
    0.75
     lashes
    0.73
     scapego
    0.71
     lash
    0.71
     blame
    0.70
     harshly
    0.67
     forgiven
    0.65
     accuse
    0.64
    Activations Density 0.012%

    No Known Activations