INDEX

Explanations

personal usernames or handles containing the word "dy"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the substring "dy" 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

insula

-0.78

ulative

-0.71

krit

-0.71

ilitarian

-0.71

itudes

-0.68

ompl

-0.66

umerable

-0.66

 shorth

-0.65

FAM

-0.65

iquid

-0.65

POSITIVE LOGITS

dy

1.06

lde

0.85

dy

0.80

wallet

0.80

Dy

0.76

fed

0.73

boys

0.72

rome

0.70

igger

0.70

marine

0.70

Activations Density 0.013%

personal usernames or handles containing the word "dy"

occurrences of the substring "dy" in various contexts

No Comments

No Known Activations