INDEX

Explanations

phrases indicating comprehension or lack of understanding about a situation or behavior

oai_token-act-pair · gpt-3.5-turbo

expressions of understanding or awareness of situations and experiences

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

yz

-0.69

 simul

-0.64

 sidx

-0.63

ixt

-0.62

odder

-0.61

 bonding

-0.60

rouse

-0.60

apego

-0.58

 contamin

-0.58

andals

-0.58

POSITIVE LOGITS

 myself

0.99

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.88

à¼

0.88

âĶĢâĶĢâĶĢâĶĢ

0.78

remember

0.68

Í

0.66

ĸ

0.66

agree

0.64

 âĶľ

0.64

my

0.60

Activations Density 0.326%

phrases indicating comprehension or lack of understanding about a situation or behavior

expressions of understanding or awareness of situations and experiences

No Comments

No Known Activations