INDEX

Explanations

occurrences of words or phrases related to continuation or persistence

oai_token-act-pair · gpt-3.5-turbo

phrases indicating continuity or persistence over time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Pacific

-0.80

 Flavoring

-0.77

umers

-0.76

externalActionCode

-0.76

ESS

-0.73

vantage

-0.73

asus

-0.71

âĸĳ

-0.70

ï¸

-0.70

Qual

-0.70

POSITIVE LOGITS

 disclaim

0.78

 accounting

0.77

 exhaustive

0.71

 caveats

0.70

 watered

0.69

 diminishing

0.68

 slight

0.67

 whiff

0.66

 declining

0.65

 acknowledging

0.64

Activations Density 0.057%

occurrences of words or phrases related to continuation or persistence

phrases indicating continuity or persistence over time

No Comments

No Known Activations