Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 0-RES-JB
    6. 7483
    Prev
    Next
    INDEX
    Explanations

    words related to lethality or dangerous situations

    oai_token-act-pair · gpt-3.5-turbo

    keywords related to fatal outcomes and dangerous situations

    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
     Sponge
    -0.76
    bt
    -0.75
    BN
    -0.74
     Tourism
    -0.69
    dates
    -0.68
     Theme
    -0.67
    bum
    -0.67
    BLIC
    -0.67
    ¹
    -0.66
    rastructure
    -0.65
    POSITIVE LOGITS
     lethal
    3.70
    lethal
    2.53
     Lethal
    1.84
     deadly
    1.68
     fatal
    1.60
     murderous
    1.24
     killing
    1.21
     dangerous
    1.21
     leth
    1.19
     toxic
    1.16
    Activations Density 0.014%

    No Known Activations