INDEX

Explanations

names of locations or institutions, specifically related to news articles

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Washington."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gger

-0.74

order

-0.74

fm

-0.70

{*

-0.66

Boo

-0.65

lass

-0.64

loss

-0.63

hist

-0.63

Occ

-0.63

eworld

-0.63

POSITIVE LOGITS

ASHINGTON

1.18

aukee

1.12

WASHINGTON

0.96

ashtra

0.95

DC

0.87

MENTS

0.85

NESS

0.84

ADA

0.83

 STATE

0.82

nesday

0.80

Activations Density 0.007%

names of locations or institutions, specifically related to news articles

occurrences of the word "Washington."

No Comments

No Known Activations