INDEX

Explanations

phrases indicating certainty or doubt

oai_token-act-pair · gpt-3.5-turbo

expressions of certainty or assumptions regarding people's knowledge or opinions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

>]

-0.73

ĻĤ

-0.72

inth

-0.64

ilaterally

-0.62

Exit

-0.62

iasco

-0.62

 effectively

-0.61

sole

-0.59

 supposedly

-0.59

ocre

-0.59

POSITIVE LOGITS

 delighted

0.80

 wouldn

0.80

 benefited

0.77

 abhor

0.73

 disappointed

0.71

 minded

0.71

 thrilled

0.70

 hated

0.69

 horr

0.69

 pissed

0.68

Activations Density 0.286%

phrases indicating certainty or doubt

expressions of certainty or assumptions regarding people's knowledge or opinions

No Comments

No Known Activations