INDEX

Explanations

measurements or numerical values expressed in meters

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter "m"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

DragonMagazine

-0.96

ãĥ´ãĤ¡

-0.86

é¾įåĸļå£«

-0.78

ãĤ¡

-0.77

gerald

-0.76

ä½ľ

-0.74

ãģį

-0.71

OPLE

-0.70

ãĥ¼ãĤ¯

-0.69

 Eleven

-0.69

POSITIVE LOGITS

igr

1.04

apping

1.03

iddling

1.02

akin

1.00

appings

1.00

apper

0.99

ixed

0.98

ichael

0.98

amm

0.95

oto

0.94

Activations Density 0.018%

measurements or numerical values expressed in meters

occurrences of the letter "m"

No Comments

No Known Activations