INDEX

Explanations

references to the word "Me" as a personal identifier

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Me."

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

paralle

-0.70

medium

-0.69

*/(

-0.69

çĶŁ

-0.68

edged

-0.68

 legality

-0.67

ordinate

-0.66

raints

-0.65

Closure

-0.65

tip

-0.62

POSITIVE LOGITS

asuring

1.02

zzo

1.01

asured

0.99

adow

0.98

asure

0.91

adows

0.90

lees

0.90

cca

0.89

ghan

0.85

aning

0.85

Activations Density 0.023%

references to the word "Me" as a personal identifier

occurrences of the word "Me."

No Comments

No Known Activations