INDEX

Explanations

phrases related to opposition or conflict

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Against."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 static

-0.62

 license

-0.61

bed

-0.61

 convey

-0.61

 pipe

-0.61

kered

-0.61

 sight

-0.60

nic

-0.60

 bright

-0.59

 basket

-0.59

POSITIVE LOGITS

 Against

3.96

Against

2.91

against

2.56

 against

1.98

AGA

1.28

 Towards

1.25

 Versus

1.09

Vs

1.01

Anti

1.00

 Anti

1.00

Activations Density 0.014%

phrases related to opposition or conflict

instances of the word "Against."

No Comments

No Known Activations