INDEX

Explanations

the word "the" followed by a verb

oai_token-act-pair · gpt-3.5-turbo

instances of the word "the" and related expressions emphasizing recognition or communication

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

arcity

-0.76

ature

-0.68

ALSE

-0.67

Â²

-0.66

iasco

-0.66

avid

-0.64

olutely

-0.63

zai

-0.63

IUM

-0.62

asionally

-0.61

POSITIVE LOGITS

 gist

1.15

job

1.08

nod

1.08

 memo

1.07

 message

1.06

 hang

1.04

 knack

0.98

 impression

0.96

 chance

0.96

 brunt

0.95

Activations Density 0.062%

the word "the" followed by a verb

instances of the word "the" and related expressions emphasizing recognition or communication

No Comments

No Known Activations