INDEX

Explanations

words containing the string "ar" with high activation values, suggesting a potential focus on names or terms containing this string

oai_token-act-pair · gpt-3.5-turbo

words or phrases related to specific geographical locations or landmarks

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĳå£«

-0.72

-+-+

-0.72

»Ĵ

-0.69

ICAN

-0.66

 Bened

-0.66

ãĥ¼ãĥĨ

-0.64

\\\\\\\\

-0.61

Ļ½

-0.60

ENA

-0.58

 Kinnikuman

-0.58

POSITIVE LOGITS

ibaba

0.95

iland

0.92

bucks

0.91

kee

0.76

buck

0.73

lde

0.72

agger

0.71

kees

0.71

interstitial

0.70

onga

0.68

Activations Density 0.040%

words containing the string "ar" with high activation values, suggesting a potential focus on names or terms containing this string

words or phrases related to specific geographical locations or landmarks

No Comments

No Known Activations