INDEX

Explanations

mentions of the word "peers"

oai_token-act-pair · gpt-3.5-turbo

references to peers and peer relationships

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

othe

-0.70

ournal

-0.69

ertility

-0.65

otide

-0.64

 Liberties

-0.63

 Tycoon

-0.62

ohyd

-0.58

 Ukrain

-0.58

ono

-0.57

oman

-0.56

POSITIVE LOGITS

 peers

1.27

hip

0.85

peer

0.84

ystem

0.79

flix

0.76

mates

0.76

hips

0.74

Reviewed

0.74

 classmates

0.72

hall

0.71

Activations Density 0.007%

mentions of the word "peers"

references to peers and peer relationships

No Comments

No Known Activations