INDEX

Explanations

years, specifically focusing on the years 1971 and 1973

oai_token-act-pair · gpt-3.5-turbo

specific years, particularly the early 1970s

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

tro

-0.84

 entirety

-0.75

edo

-0.71

cho

-0.71

unders

-0.70

unw

-0.68

plan

-0.68

Edge

-0.68

unin

-0.67

vere

-0.65

POSITIVE LOGITS

ĸļ

0.91

 onwards

0.84

å¹

0.74

 onward

0.74

0.73

0.69

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.68

0.66

 Chevrolet

0.65

Activations Density 0.025%

years, specifically focusing on the years 1971 and 1973

specific years, particularly the early 1970s

No Comments

No Known Activations