INDEX

Explanations

the word "that"

oai_token-act-pair · gpt-3.5-turbo

conjunctions and phrases that introduce assertions or evidence within a statement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rete

-0.75

fect

-0.70

oleon

-0.68

oret

-0.68

icking

-0.68

Course

-0.68

nec

-0.67

aler

-0.67

NES

-0.67

oco

-0.67

POSITIVE LOGITS

 contradicts

1.38

 accompanies

1.33

 proves

1.33

 justifies

1.31

 suggests

1.31

 undermines

1.26

 indicates

1.22

 confirms

1.16

 distinguishes

1.15

 implies

1.15

Activations Density 0.163%

the word "that"

conjunctions and phrases that introduce assertions or evidence within a statement

No Comments

No Known Activations