INDEX

Explanations

adjectives related to intensity or importance

oai_token-act-pair · gpt-3.5-turbo

intensifiers that emphasize qualities or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Palest

-0.66

bright

-0.64

onto

-0.60

ee

-0.60

Cle

-0.60

CLE

-0.59

cour

-0.59

Spr

-0.59

ãĥĩ

-0.59

Tw

-0.59

POSITIVE LOGITS

(>

0.83

 indeed

0.71

xual

0.64

idious

0.61

Ö

0.60

illac

0.60

WIN

0.59

velength

0.59

ensable

0.59

 âĶľ

0.58

Activations Density 0.174%

adjectives related to intensity or importance

intensifiers that emphasize qualities or conditions

No Comments

No Known Activations