INDEX

Explanations

prefixes or suffixes in words

oai_token-act-pair · gpt-3.5-turbo

prefixes and word parts commonly associated with a legal or governance context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

©¶æ¥µ

-0.79

 Wilde

-0.69

 Elves

-0.67

 Immunity

-0.65

HAEL

-0.63

WOOD

-0.62

BALL

-0.62

 Aires

-0.62

è£ħ

-0.61

 unsur

-0.60

POSITIVE LOGITS

ruction

1.13

tenance

0.98

acters

0.98

gment

0.97

anced

0.97

ighed

0.92

vised

0.91

ctions

0.91

ction

0.91

lected

0.91

Activations Density 0.133%

prefixes or suffixes in words

prefixes and word parts commonly associated with a legal or governance context

No Comments

No Known Activations