INDEX

Explanations

mentions of government entities

oai_token-act-pair · gpt-3.5-turbo

references to governmental entities and their actions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dimension

-0.71

Dialog

-0.70

 Rider

-0.69

Magikarp

-0.68

cases

-0.67

Accessory

-0.65

 Surge

-0.64

 Virtue

-0.62

Charges

-0.62

 Virus

-0.62

POSITIVE LOGITS

chool

1.06

 subsidized

0.85

 governments

0.79

 decree

0.77

 funded

0.75

 allied

0.74

wide

0.72

 sanctioned

0.72

 appropriated

0.71

 bureaucr

0.71

Activations Density 0.025%

mentions of government entities

references to governmental entities and their actions

No Comments

No Known Activations