INDEX

Explanations

references to specific durations of time

oai_token-act-pair · gpt-3.5-turbo

time durations and intervals

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

egu

-0.77

 Advis

-0.62

kinson

-0.62

guyen

-0.61

Redd

-0.59

acus

-0.59

eller

-0.58

eele

-0.58

 Redmond

-0.57

pend

-0.57

POSITIVE LOGITS

 straight

1.23

 afterward

1.10

 uninterrupted

1.07

 afterwards

1.01

 apiece

0.95

 consecut

0.95

 longer

0.93

straight

0.91

 before

0.88

 thereafter

0.87

Activations Density 0.117%

references to specific durations of time

time durations and intervals

No Comments

No Known Activations