INDEX

Explanations

items listed in a structured format containing a mix of varied information

oai_token-act-pair · gpt-3.5-turbo

lists and comparisons related to demographics and societal observations

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

unin

-0.54

 insepar

-0.54

egu

-0.52

ĺħ

-0.51

 Powered

-0.51

¬¼

-0.51

 slightest

-0.51

ocument

-0.50

odor

-0.50

escap

-0.50

POSITIVE LOGITS

 than

2.87

than

2.63

 Than

2.18

 compared

1.15

TH

0.91

Th

0.84

 rather

0.81

 then

0.80

worldly

0.79

 THEN

0.75

Activations Density 1.491%

items listed in a structured format containing a mix of varied information

lists and comparisons related to demographics and societal observations

No Comments

No Known Activations