INDEX

Explanations

phrases related to the speaker expressing knowledge or understanding

oai_token-act-pair · gpt-3.5-turbo

expressions of certainty or knowledge

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

isco

-0.86

 sidx

-0.83

acies

-0.80

onding

-0.80

taboola

-0.77

pload

-0.73

cific

-0.72

aukee

-0.71

issance

-0.70

emetery

-0.70

POSITIVE LOGITS

 firsthand

0.90

 nothing

0.77

ledged

0.76

ledge

0.75

how

0.75

nothing

0.70

 beforehand

0.67

 what

0.66

 exactly

0.64

 personally

0.63

Activations Density 0.046%

phrases related to the speaker expressing knowledge or understanding

expressions of certainty or knowledge

No Comments

No Known Activations