INDEX

Explanations

phrases related to chronological events or timelines

oai_token-act-pair · gpt-3.5-turbo

the concept of series or sequences in relation to content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

egu

-0.76

Test

-0.67

 externalToEVAOnly

-0.66

 RELE

-0.65

 OFFIC

-0.62

 este

-0.62

 CONTIN

-0.62

redits

-0.61

sob

-0.61

Pwr

-0.61

POSITIVE LOGITS

emouth

0.75

bole

0.70

naissance

0.67

²¾

0.67

conservancy

0.65

 leisure

0.65

vantage

0.64

purpose

0.63

coni

0.61

ork

0.60

Activations Density 0.000%

phrases related to chronological events or timelines

the concept of series or sequences in relation to content

No Comments

No Known Activations