INDEX

Explanations

superlative words, particularly ones emphasizing importance or intensity

oai_token-act-pair · gpt-3.5-turbo

the word "most" in various contexts indicating emphasis or degree

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

anyl

-0.65

instead

-0.64

 Films

-0.61

 Stead

-0.60

heid

-0.57

 Creed

-0.57

IPS

-0.57

 Chips

-0.57

 Chains

-0.56

onut

-0.56

POSITIVE LOGITS

 importantly

1.18

 likely

0.94

likely

0.93

 assured

0.85

 prominently

0.84

 important

0.82

afa

0.79

 notably

0.79

rar

0.77

 egreg

0.76

Activations Density 0.064%

superlative words, particularly ones emphasizing importance or intensity

the word "most" in various contexts indicating emphasis or degree

No Comments

No Known Activations