INDEX

Explanations

personal pronouns followed by some form of "be" verb

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "it" and variations of the term "to be"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 PLUS

-0.91

fleet

-0.78

 srfAttach

-0.73

INS

-0.72

ovo

-0.70

.?

-0.67

ļéĨĴ

-0.67

cffff

-0.64

 GOODMAN

-0.64

.ï¿½

-0.63

POSITIVE LOGITS

 technically

1.14

 initially

1.02

 admittedly

0.98

 laud

0.98

 undoubtedly

0.97

may

0.94

 ostensibly

0.93

 certainly

0.91

 theoretically

0.88

 sounds

0.87

Activations Density 0.283%

personal pronouns followed by some form of "be" verb

phrases that include the word "it" and variations of the term "to be"

No Comments

No Known Activations