INDEX

Explanations

pronouns followed by a verb indicating an action

oai_token-act-pair · gpt-3.5-turbo

repeated references to the word "it."

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

Priv

-0.69

quartered

-0.63

holding

-0.62

anton

-0.60

 Spending

-0.60

behind

-0.60

dding

-0.58

Jaw

-0.58

telling

-0.58

warning

-0.57

POSITIVE LOGITS

 disappears

1.16

 happens

1.15

 happened

1.14

chy

1.12

alian

1.12

 rains

1.11

unes

1.10

 occurs

1.06

 becomes

1.06

 fades

1.03

Activations Density 0.326%

pronouns followed by a verb indicating an action

repeated references to the word "it."

No Comments

No Known Activations