INDEX

Explanations

adjectives describing a particularly surprising or impressive quality

oai_token-act-pair · gpt-3.5-turbo

words that convey surprise or awe

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Mos

-0.70

à

-0.68

utor

-0.68

Neh

-0.67

ado

-0.67

rists

-0.65

mable

-0.65

Default

-0.64

arians

-0.64

malink

-0.63

POSITIVE LOGITS

 curing

0.66

 rover

0.65

 Hercules

0.62

 centrif

0.61

Dri

0.61

WD

0.61

 rejuven

0.61

clud

0.61

 fading

0.60

 lunar

0.59

Activations Density 0.000%

adjectives describing a particularly surprising or impressive quality

words that convey surprise or awe

No Comments

No Known Activations