INDEX

Explanations

computer-related specific steps or instructions

oai_token-act-pair · gpt-3.5-turbo

empty placeholder tokens or formatting elements likely indicating structural components of the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

kef

-0.70

.","

-0.68

phia

-0.67

tin

-0.66

itent

-0.61

hung

-0.61

bilt

-0.60

''

-0.59

 Antar

-0.58

cle

-0.58

POSITIVE LOGITS

jamin

0.94

resa

0.77

iculty

0.77

withstanding

0.72

 Sample

0.70

ordable

0.69

bidden

0.69

ertodd

0.68

 Contents

0.67

llor

0.66

Activations Density 0.181%

computer-related specific steps or instructions

empty placeholder tokens or formatting elements likely indicating structural components of the text

No Comments

No Known Activations