INDEX

Explanations

phrases related to fulfilling promises or expectations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "delivering."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

oos

-0.69

wikipedia

-0.68

rawn

-0.68

 Reserved

-0.61

obia

-0.61

esian

-0.61

oor

-0.61

 FANTASY

-0.60

wik

-0.60

wiki

-0.59

POSITIVE LOGITS

 delivering

3.27

 deliver

2.50

 delivers

2.21

 delivered

2.12

 delivery

2.08

 Deliver

1.83

 Delivery

1.69

 deliveries

1.68

Delivery

1.60

 distributing

1.25

Activations Density 0.017%

phrases related to fulfilling promises or expectations

occurrences of the word "delivering."

No Comments

No Known Activations