INDEX

Explanations

phrases related to policies and government decisions

oai_token-act-pair · gpt-3.5-turbo

references to "policy" and its related contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

•Art-related Features

•newlist

No Comments

Negative Logits

 Stain

-0.78

DAY

-0.74

 Sparkle

-0.71

 Norn

-0.69

 Templ

-0.69

igans

-0.68

Ñĭ

-0.66

lighting

-0.65

 Takeru

-0.64

 Warranty

-0.63

POSITIVE LOGITS

making

1.08

 makers

0.98

 prescriptions

0.98

makers

0.92

 agenda

0.89

 objectives

0.88

 stances

0.88

 correctness

0.88

 advisers

0.86

 advisors

0.86

Activations Density 0.047%

phrases related to policies and government decisions

references to "policy" and its related contexts

No Comments

No Known Activations