INDEX

Explanations

occurrences of the word "Luckily."

oai_token-act-pair · gpt-3.5-turbo

expressions of positive sentiment or good fortune

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

wa

-0.68

 endorsement

-0.66

 dual

-0.63

 parallel

-0.63

 greater

-0.59

 voting

-0.59

 bias

-0.59

 ment

-0.58

 closed

-0.58

 supporting

-0.58

POSITIVE LOGITS

 Luckily

3.21

 Fortunately

3.16

 Thankfully

3.07

Luckily

2.96

Fortunately

2.86

Thankfully

2.77

 luckily

2.35

 fortunately

2.16

 thankfully

1.82

 Hopefully

1.49

Activations Density 0.023%

occurrences of the word "Luckily."

expressions of positive sentiment or good fortune

No Comments

No Known Activations