INDEX

Explanations

dates written in a specific style

oai_token-act-pair · gpt-3.5-turbo

dates, specifically those in July

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

ngth

-0.85

hyde

-0.79

 Reincarn

-0.76

xual

-0.76

cipline

-0.72

 Sight

-0.70

culosis

-0.70

anwhile

-0.68

astern

-0.68

hower

-0.66

POSITIVE LOGITS

ienne

1.07

iana

1.06

iets

0.98

iano

0.96

iom

0.95

iane

0.94

ani

0.92

arro

0.90

ians

0.90

iani

0.87

Activations Density 0.009%

dates written in a specific style

dates, specifically those in July

No Comments

No Known Activations