INDEX

Explanations

references to specific locations or names, especially related to current events or politics

oai_token-act-pair · gpt-3.5-turbo

instances of the word "the."

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

aba

-0.77

 because

-0.73

thood

-0.72

elaide

-0.71

eno

-0.71

Ò

-0.70

acy

-0.68

!!!!

-0.67

leground

-0.67

ional

-0.65

POSITIVE LOGITS

 slightest

1.04

 simplest

1.04

 biggest

1.03

 vast

1.01

odore

1.01

 latter

1.00

resa

0.99

 majority

0.99

 entire

0.98

 oldest

0.98

Activations Density 0.274%

references to specific locations or names, especially related to current events or politics

instances of the word "the."

No Comments

No Known Activations