© 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. 0-RES-JB
    11. 16814
    Prev
    Next
    INDEX
    Explanations

    references to various types of speeches

    oai_token-act-pair · gpt-3.5-turbo

    references to formal addresses or public speaking events

    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
    Tank
    -0.67
    finder
    -0.66
    omi
    -0.65
     Tank
    -0.65
     Warehouse
    -0.64
    arant
    -0.63
    Ãĥ
    -0.61
     Skydragon
    -0.60
     Jagu
    -0.60
     TTL
    -0.60
    POSITIVE LOGITS
     speech
    3.84
     Speech
    3.19
    speech
    3.17
     speeches
    2.19
     sermon
    1.38
     keynote
    1.35
     lecture
    1.25
     speakers
    1.23
     speaking
    1.22
     discourse
    1.22
    Activations Density 0.018%

    No Known Activations