INDEX

Explanations

words and phrases that end in "ly."

oai_token-act-pair · gpt-3.5-turbo

adverbs that modify the intensity or manner of actions or descriptors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 GOODMAN

-0.93

ilater

-0.90

afety

-0.81

eport

-0.79

emetery

-0.71

ERY

-0.71

ĸļ

-0.70

ifacts

-0.69

irtual

-0.68

natureconservancy

-0.68

POSITIVE LOGITS

present

0.86

adv

0.84

rics

0.77

 discharged

0.76

 priced

0.75

tics

0.74

ffe

0.73

alpha

0.71

gged

0.71

puff

0.71

Activations Density 0.036%

words and phrases that end in "ly."

adverbs that modify the intensity or manner of actions or descriptors

No Comments

No Known Activations