INDEX

Explanations

questions

oai_token-act-pair · gpt-3.5-turbo

questions in the text

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

 corrid

-0.76

 swamp

-0.67

lock

-0.66

 inclusive

-0.65

 portal

-0.65

ikuman

-0.64

agne

-0.64

éĹĺ

-0.64

der

-0.64

 booked

-0.63

POSITIVE LOGITS

 Nope

1.37

 Absolutely

1.31

 Probably

1.29

 Certainly

1.29

 Possibly

1.28

 Maybe

1.14

 Perhaps

1.13

 Does

1.13

Yes

1.13

 Surely

1.12

Activations Density 0.090%

questions

questions in the text

No Comments

No Known Activations