© 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. 3-RES-JB
    11. 22587
    Prev
    Next
    INDEX
    Explanations

    descriptions of things or situations by using verbs like "described" and various adjectives

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "describe" or its variants

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ammy
    -0.76
    aghetti
    -0.75
    assi
    -0.72
    alos
    -0.70
    umo
    -0.68
    iasm
    -0.66
    app
    -0.66
    ffic
    -0.66
    ublic
    -0.65
    sei
    -0.62
    POSITIVE LOGITS
     descriptions
    0.86
    REDACTED
    0.83
    urated
    0.76
     aloud
    0.76
    urally
    0.74
    %]
    0.74
    ounces
    0.72
     describ
    0.72
     symptoms
    0.69
    Filename
    0.68
    Activations Density 0.026%

    No Known Activations