INDEX

Explanations

instances where something is significantly more/less or better/worse than something else

oai_token-act-pair · gpt-3.5-turbo

the word "significantly" and related phrases indicating substantial changes or impacts

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

tein

-0.86

ifully

-0.77

elf

-0.76

nar

-0.75

eer

-0.74

iens

-0.72

elsen

-0.70

olan

-0.68

ies

-0.68

icular

-0.66

POSITIVE LOGITS

 impacting

0.93

 effected

0.80

 deterior

0.78

 impacted

0.77

 outper

0.75

 benefited

0.75

 influenced

0.74

 degrade

0.74

 proport

0.73

significant

0.73

Activations Density 0.012%

instances where something is significantly more/less or better/worse than something else

the word "significantly" and related phrases indicating substantial changes or impacts

No Comments

No Known Activations