INDEX

Explanations

references to a specific retail store chain

oai_token-act-pair · gpt-3.5-turbo

mentions of a specific retail brand

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Tayyip

-0.72

 deaf

-0.67

UTH

-0.67

icter

-0.65

 untreated

-0.64

 fertile

-0.63

HEAD

-0.62

 polarization

-0.62

 pregnant

-0.61

 Anchorage

-0.61

POSITIVE LOGITS

mart

1.18

inez

0.97

ards

0.96

Mart

0.95

ukong

0.93

inho

0.91

 Stores

0.87

ertodd

0.87

inet

0.87

lain

0.86

Activations Density 0.006%

references to a specific retail store chain

mentions of a specific retail brand

No Comments

No Known Activations