INDEX

Explanations

question marks

oai_token-act-pair · gpt-3.5-turbo

numerous question marks indicating inquiries or uncertainty

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Nadu

-0.74

apers

-0.74

shaw

-0.70

 slam

-0.65

ality

-0.64

 Swanson

-0.63

 helicop

-0.61

 marsh

-0.61

 tele

-0.59

 encount

-0.59

POSITIVE LOGITS

ï¿½ï¿½ï¿½ï¿½

1.03

utm

1.01

?:

0.96

.?

0.94

 Nope

0.82

crop

0.77

/?

0.76

ï¸ı

0.75

rencies

0.74

Huh

0.73

Activations Density 0.059%

question marks

numerous question marks indicating inquiries or uncertainty

No Comments

No Known Activations