© 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. 8-RES-JB
    11. 20340
    Prev
    Next
    INDEX
    Explanations

    contrasting statements, particularly focusing on disproving or negating an initial idea

    oai_token-act-pair · gpt-3.5-turbo

    contrasting statements that introduce clarification or correction

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    asks
    -0.71
    illary
    -0.66
    meta
    -0.64
    ILLE
    -0.63
    nat
    -0.62
    ct
    -0.62
    holder
    -0.61
    enter
    -0.60
    orter
    -0.60
    ory
    -0.59
    POSITIVE LOGITS
     rather
    1.38
     nevertheless
    1.12
     merely
    1.10
    rather
    1.09
     nonetheless
    1.02
     suffice
    0.98
     Rather
    0.96
     instead
    0.91
     simply
    0.90
     luckily
    0.84
    Activations Density 0.110%

    No Known Activations