© 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. 11-RES-JB
    11. 16599
    Prev
    Next
    INDEX
    Explanations

    statements related to posing threats or dangers

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating threats or dangers

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    endars
    -0.75
     Journals
    -0.74
     Franks
    -0.71
    essen
    -0.67
     Celebr
    -0.65
     tunes
    -0.65
    aturday
    -0.63
     Manufact
    -0.63
     desserts
    -0.62
    utenberg
    -0.62
    POSITIVE LOGITS
     threat
    1.58
     hazard
    1.46
     danger
    1.42
     risk
    1.42
    threat
    1.30
     risks
    1.15
     challenge
    1.12
     existential
    1.09
     peril
    1.09
     hazards
    1.07
    Activations Density 0.121%

    No Known Activations