INDEX

Explanations

action verbs related to observing or analyzing something within a text

oai_token-act-pair · gpt-3.5-turbo

words indicating observation or discovery

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

)]

-0.76

lic

-0.63

ICO

-0.62

xon

-0.62

 livest

-0.60

phrine

-0.58

imeters

-0.57

nor

-0.56

ijah

-0.55

--+

-0.55

POSITIVE LOGITS

 that

1.04

why

1.03

 exactly

0.89

how

0.86

 something

0.86

 plenty

0.86

 unmist

0.86

 quite

0.84

 firsthand

0.83

WHY

0.82

Activations Density 0.188%

action verbs related to observing or analyzing something within a text

words indicating observation or discovery

No Comments

No Known Activations