INDEX

Explanations

capitalized acronyms ending with 'OT'

oai_token-act-pair · gpt-3.5-turbo

occurrences of the abbreviation "OT"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

uckland

-0.74

 fixme

-0.71

bler

-0.67

kers

-0.67

ãĥĭ

-0.65

ieties

-0.65

antha

-0.61

 sinks

-0.61

issance

-0.60

worms

-0.60

POSITIVE LOGITS

assium

1.09

OT

1.04

TL

0.99

OGR

0.99

YP

0.96

TE

0.95

ALS

0.89

TO

0.89

ECH

0.89

IC

0.88

Activations Density 0.008%

capitalized acronyms ending with 'OT'

occurrences of the abbreviation "OT"

No Comments

No Known Activations