INDEX

Explanations

proper nouns or names, specifically focused on the name "Watkins" and "Thornhill"

oai_token-act-pair · gpt-3.5-turbo

mentions of specific athletes, particularly focusing on names associated with sports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

iltration

-0.70

asm

-0.65

arily

-0.64

onyms

-0.61

hal

-0.61

 prisoner

-0.61

Occup

-0.60

erity

-0.60

liber

-0.60

cffffcc

-0.59

POSITIVE LOGITS

 Watkins

1.02

 Beir

0.82

combe

0.75

folios

0.74

creen

0.73

 Thorn

0.73

apons

0.73

ipers

0.70

yth

0.69

wana

0.69

Activations Density 0.008%

proper nouns or names, specifically focused on the name "Watkins" and "Thornhill"

mentions of specific athletes, particularly focusing on names associated with sports

No Comments

No Known Activations