INDEX

Explanations

references to specific entities or organizations mentioned in a context of news or events

oai_token-act-pair · gpt-3.5-turbo

instances of the word "the"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

thood

-0.78

RIC

-0.77

lement

-0.73

beard

-0.73

bear

-0.71

aurus

-0.69

Iterator

-0.68

lessly

-0.68

upon

-0.67

ties

-0.66

POSITIVE LOGITS

 periphery

1.22

 basis

1.19

 occasion

1.17

 heels

1.16

 same

1.14

 sidelines

1.14

 brink

1.10

 verge

1.06

 forefront

1.04

 premise

1.04

Activations Density 0.223%

references to specific entities or organizations mentioned in a context of news or events

instances of the word "the"

No Comments

No Known Activations