INDEX

Explanations

words related to things that are hard or challenging

oai_token-act-pair · gpt-3.5-turbo

words with prefixes indicating negation or lack of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

aurus

-0.76

 Origins

-0.70

 Lauder

-0.68

auga

-0.67

 Dynamics

-0.66

 impressions

-0.64

eur

-0.64

 Carbuncle

-0.63

Ti

-0.63

 Rams

-0.62

POSITIVE LOGITS

ported

1.08

ipolar

0.98

romising

0.97

ainted

0.96

leased

0.95

ishable

0.94

stretched

0.93

modified

0.93

ublic

0.92

ended

0.90

Activations Density 0.027%

words related to things that are hard or challenging

words with prefixes indicating negation or lack of something

No Comments

No Known Activations