INDEX

Explanations

references to the word "telegraph"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Telegraph" and its variants

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

regn

-0.72

rolet

-0.69

 awoken

-0.63

MLG

-0.62

 Louie

-0.62

atri

-0.60

jury

-0.59

tesy

-0.59

 Williamson

-0.59

amac

-0.58

POSITIVE LOGITS

legraph

1.10

 Tycoon

0.85

ic

0.84

shire

0.81

orial

0.80

biz

0.80

soever

0.78

gow

0.77

igel

0.76

icate

0.75

Activations Density 0.022%

references to the word "telegraph"

occurrences of the word "Telegraph" and its variants

No Comments

No Known Activations