© 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. 8-RES-JB
    11. 2142
    Prev
    Next
    INDEX
    Explanations

    mentions of extreme circumstances or events

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "extreme" in various contexts, particularly related to violence and other intense situations

    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
    athan
    -0.84
    yard
    -0.81
    icative
    -0.76
    glas
    -0.75
    edin
    -0.75
    bats
    -0.75
    ËĪ
    -0.73
    fman
    -0.73
    icating
    -0.72
    icate
    -0.72
    POSITIVE LOGITS
     poverty
    0.88
     vetting
    0.86
     lengths
    0.83
     temperatures
    0.79
     rendition
    0.79
     rarity
    0.78
     weather
    0.76
     amounts
    0.74
     sensitivity
    0.74
     measures
    0.74
    Activations Density 0.039%

    No Known Activations