INDEX

Explanations

negative or contrasting statements

oai_token-act-pair · gpt-3.5-turbo

negations or phrases that express inadequacy or dissatisfaction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

aughs

-0.81

Ĥİ

-0.74

etimes

-0.73

TIME

-0.72

States

-0.67

 Companies

-0.65

WAY

-0.65

 Treatment

-0.65

now

-0.64

nton

-0.64

POSITIVE LOGITS

 flashy

1.21

 overly

1.19

 terribly

1.14

 necessarily

1.12

icable

1.11

icably

1.06

epad

1.05

 overpower

1.01

 uncommon

0.95

 fancy

0.94

Activations Density 0.171%

negative or contrasting statements

negations or phrases that express inadequacy or dissatisfaction

No Comments

No Known Activations