INDEX

Explanations

negative or critical comments

oai_token-act-pair · gpt-3.5-turbo

the presence of emphatic or assertive statements 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

-0.76

handled

-0.71

quartered

-0.70

lees

-0.65

angan

-0.63

 Virtue

-0.63

rehend

-0.63

²¾

-0.62

soever

-0.62

 Result

-0.60

POSITIVE LOGITS

 forgot

0.70

 congr

0.67

 spoilers

0.64

 kidding

0.63

ain

0.62

 lied

0.62

Naw

0.60

 lousy

0.58

 Dull

0.57

 insensitive

0.57

Activations Density 0.341%

negative or critical comments

the presence of emphatic or assertive statements in the text

No Comments

No Known Activations