INDEX

Explanations

instances where a statement or story contradicts initial reports or recorded views

oai_token-act-pair · gpt-3.5-turbo

the word "that" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

atre

-0.67

stead

-0.67

Laughs

-0.65

å¸

-0.63

gur

-0.63

esi

-0.60

Twe

-0.60

DER

-0.59

kamp

-0.58

hare

-0.58

POSITIVE LOGITS

 accompanies

1.13

 prevailed

1.02

 resulted

1.00

 contradicts

0.99

 preceded

0.99

 existed

0.98

 governs

0.97

 justifies

0.95

 arose

0.95

 occurred

0.93

Activations Density 0.214%

instances where a statement or story contradicts initial reports or recorded views

the word "that" in various contexts

No Comments

No Known Activations