INDEX

Explanations

words related to specific locations, particularly related to forest areas

oai_token-act-pair · gpt-3.5-turbo

references to forests or forest-related entities

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

esthetic

-0.71

soever

-0.69

verages

-0.69

atform

-0.66

epad

-0.64

 contagious

-0.63

Magikarp

-0.63

ocial

-0.62

ifice

-0.61

demand

-0.61

POSITIVE LOGITS

 Forest

0.94

lake

0.93

 Lawn

0.93

lands

0.92

 Hills

0.86

banks

0.86

land

0.84

bank

0.78

ieri

0.78

 Stew

0.77

Activations Density 0.013%

words related to specific locations, particularly related to forest areas

references to forests or forest-related entities

No Comments

No Known Activations