INDEX

Explanations

phrases indicating a specific quantity or number

oai_token-act-pair · gpt-3.5-turbo

phrases indicating quantities or amounts, particularly "more" and "less."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

culosis

-0.69

Ney

-0.67

crime

-0.63

 Hamm

-0.62

cookie

-0.62

 Clarks

-0.61

gers

-0.61

Ike

-0.60

english

-0.60

Agent

-0.59

POSITIVE LOGITS

agonal

0.75

 equivalent

0.71

igree

0.71

inguished

0.70

inction

0.70

>]

0.68

 comparable

0.68

 versa

0.68

 beaten

0.66

NM

0.66

Activations Density 0.060%

phrases indicating a specific quantity or number

phrases indicating quantities or amounts, particularly "more" and "less."

No Comments

No Known Activations