INDEX

Explanations

action phrases related to behavior and discrepancies between stated intentions and actions

oai_token-act-pair · gpt-3.5-turbo

phrases related to hypocrisy and contradictions in behavior or statements

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

obos

-0.60

 tyres

-0.56

 pigeon

-0.54

 goodwill

-0.52

 fatigue

-0.52

 mushrooms

-0.52

oat

-0.50

 quieter

-0.50

 Jericho

-0.49

PSP

-0.49

POSITIVE LOGITS

âĢ

1.47

 âĢ

1.18

.ï¿½

1.17

.</

1.14

.*

1.13

!.

1.10

.<

1.05

1.03

*.

1.03

.(

1.03

Activations Density 0.813%

action phrases related to behavior and discrepancies between stated intentions and actions

phrases related to hypocrisy and contradictions in behavior or statements

No Comments

No Known Activations