© 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. 1-RES-JB
    11. 12310
    Prev
    Next
    INDEX
    Explanations

    information-related words, especially related to technology, news, and data analysis

    oai_token-act-pair · gpt-3.5-turbo

    references to "Information" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    kus
    -0.80
     sunny
    -0.76
     tranquil
    -0.73
     resort
    -0.71
    steen
    -0.69
    arb
    -0.68
     bed
    -0.65
    lasting
    -0.64
    days
    -0.64
    els
    -0.63
    POSITIVE LOGITS
     Information
    1.20
     Agency
    1.00
    Information
    0.98
     Awareness
    0.90
     Disclosure
    0.86
    information
    0.85
     dissemination
    0.84
     newsp
    0.82
     Availability
    0.81
     information
    0.81
    Activations Density 0.009%

    No Known Activations