INDEX

Explanations

adjectives or phrases describing high levels of intensity or importance

oai_token-act-pair · gpt-3.5-turbo

the symbol "ł" and its variations, indicating a focus on specific characters rather than content related to thematic elements

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

matic

-0.91

urated

-0.90

 primates

-0.80

raints

-0.78

 Instr

-0.76

utra

-0.73

ateurs

-0.72

ulators

-0.71

pointers

-0.71

writers

-0.70

POSITIVE LOGITS

âĶĢâĶĢ

1.32

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

1.14

âĶĢâĶĢâĶĢâĶĢ

1.03

ï¸ı

1.02

Ķ

0.95

ĺ

0.91

::::::::

0.91

à©

0.91

×Ķ

0.88

İ

0.86

Activations Density 0.127%

adjectives or phrases describing high levels of intensity or importance

the symbol "ł" and its variations, indicating a focus on specific characters rather than content related to thematic elements

No Comments

No Known Activations