INDEX

Explanations

mentions of specific organizations or events with uppercase letter acronyms

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 'E' in various contexts

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

ifier

-0.82

ifiers

-0.75

ãĤ¤ãĥĪ

-0.68

 Kenobi

-0.67

ifications

-0.64

ification

-0.64

mallow

-0.63

wagen

-0.62

MSM

-0.61

animous

-0.60

POSITIVE LOGITS

LECT

1.22

lements

1.15

ASY

1.04

ITH

1.03

instein

1.03

nerg

1.01

ighty

1.01

agles

1.00

VE

0.98

lev

0.97

Activations Density 0.029%

mentions of specific organizations or events with uppercase letter acronyms

occurrences of the letter 'E' in various contexts

No Comments

No Known Activations