INDEX

Explanations

words related to software or technical issues, specifically bugs

oai_token-act-pair · gpt-3.5-turbo

mentions of bugs, particularly in a software or gaming context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uclear

-0.84

xit

-0.80

ometown

-0.79

uther

-0.77

ACP

-0.74

asp

-0.74

estyle

-0.73

minist

-0.72

shr

-0.71

agall

-0.71

POSITIVE LOGITS

 bugs

1.29

 Bugs

1.10

bugs

1.04

 Bunny

1.00

bug

0.92

 patched

0.91

 testers

0.86

bug

0.85

 glitches

0.84

 crawling

0.83

Activations Density 0.009%

words related to software or technical issues, specifically bugs

mentions of bugs, particularly in a software or gaming context

No Comments

No Known Activations