INDEX

Explanations

recipes or specific instructions

oai_token-act-pair · gpt-3.5-turbo

references to singular or notable elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

osponsors

-0.87

etz

-0.86

ruary

-0.85

lations

-0.83

its

-0.80

loo

-0.79

Ñı

-0.79

untu

-0.77

senal

-0.76

Bay

-0.76

POSITIVE LOGITS

 thing

1.34

 exception

1.16

 caveat

1.15

 overriding

1.07

 glaring

1.04

 overarching

1.03

 downside

1.00

 flaw

1.00

 drawback

1.00

 aspect

0.99

Activations Density 0.123%

recipes or specific instructions

references to singular or notable elements

No Comments

No Known Activations