INDEX

Explanations

words related to the term "Borg"

oai_token-act-pair · gpt-3.5-turbo

references to the character "Borg" from the Star Trek franchise

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

IELD

-0.78

lled

-0.64

perature

-0.62

++++++++++++++++

-0.59

PLIED

-0.59

 Eliot

-0.59

 Appalachian

-0.59

 grades

-0.58

zanne

-0.57

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.57

POSITIVE LOGITS

 Borg

1.15

wana

0.99

hetti

0.91

uno

0.88

ulla

0.87

oreal

0.82

olini

0.80

haus

0.80

hett

0.79

leneck

0.78

Activations Density 0.005%

words related to the term "Borg"

references to the character "Borg" from the Star Trek franchise

No Comments

No Known Activations