© 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. 17094
    Prev
    Next
    INDEX
    Explanations

    phrases related to economic or financial contexts

    oai_token-act-pair · gpt-3.5-turbo

    expressions conveying concern or impending negative outcomes

    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
    onom
    -0.64
     Pog
    -0.62
    natureconservancy
    -0.53
    iple
    -0.53
     Luk
    -0.53
    ipeg
    -0.53
     Redmond
    -0.52
     Coyotes
    -0.51
     Flyers
    -0.49
    OUP
    -0.48
    POSITIVE LOGITS
    anwhile
    0.92
     respectively
    0.89
     thereby
    0.81
    }.
    0.70
     consequently
    0.64
     culminated
    0.63
     culminating
    0.63
    )).
    0.62
    thus
    0.61
    ]."
    0.60
    Activations Density 1.822%

    No Known Activations