INDEX

Explanations

the word "like."

oai_token-act-pair · gpt-3.5-turbo

the phrase "if you like" in various contexts

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

inas

-0.81

arta

-0.76

esi

-0.74

itta

-0.71

chin

-0.71

 enthusi

-0.71

cci

-0.70

anthrop

-0.69

isman

-0.69

ospel

-0.67

POSITIVE LOGITS

lihood

1.43

lier

0.99

liest

0.94

ably

0.94

 ours

0.80

liness

0.79

minded

0.78

 yours

0.72

 liking

0.71

0.69

Activations Density 0.061%

the word "like."

the phrase "if you like" in various contexts

No Comments

No Known Activations