INDEX

Explanations

mentions of bodies of water, specifically lakes

oai_token-act-pair · gpt-3.5-turbo

references to lakes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ulate

-0.70

AAAAAAAA

-0.69

edy

-0.68

nir

-0.67

eon

-0.66

STEM

-0.64

ahon

-0.63

 tremend

-0.63

cul

-0.62

ulously

-0.61

POSITIVE LOGITS

 lake

1.19

front

1.11

 trout

1.01

 basin

1.00

 lakes

0.98

 reservoir

0.93

 Erie

0.89

 Rapids

0.83

ishly

0.82

urst

0.82

Activations Density 0.007%

mentions of bodies of water, specifically lakes

references to lakes

No Comments

No Known Activations