INDEX

Explanations

words related to articles, information, directions, instructions, and references

oai_token-act-pair · gpt-3.5-turbo

mentions of information or support resources

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

berries

-0.74

gd

-0.74

gger

-0.67

etheless

-0.65

bey

-0.63

bay

-0.60

egu

-0.60

å£«

-0.59

medi

-0.59

lly

-0.59

POSITIVE LOGITS

 sake

1.01

 purposes

0.97

 refres

0.87

inkle

0.73

 reasons

0.71

icion

0.70

ummies

0.65

 adventurous

0.65

teness

0.62

 reprint

0.62

Activations Density 0.121%

words related to articles, information, directions, instructions, and references

mentions of information or support resources

No Comments

No Known Activations