INDEX

Explanations

proper nouns related to people or places

oai_token-act-pair · gpt-3.5-turbo

instances of the term "ra" across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ĳå£«

-0.77

 GOODMAN

-0.75

iaries

-0.69

é¾

-0.69

lace

-0.65

 curfew

-0.65

regor

-0.63

OW

-0.62

 charism

-0.62

erm

-0.62

POSITIVE LOGITS

irie

1.16

ven

1.16

fters

1.14

fter

1.13

ppy

1.09

eus

1.06

ving

1.05

plets

1.01

quez

0.99

zzo

0.99

Activations Density 0.017%

proper nouns related to people or places

instances of the term "ra" across various contexts

No Comments

No Known Activations