INDEX

Explanations

references to specific locations or places, particularly those containing the word "wood"

oai_token-act-pair · gpt-3.5-turbo

references to locations or entities that include the word "wood."

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

 polar

-0.84

utations

-0.71

xen

-0.64

 anat

-0.62

 uncomp

-0.62

 kindred

-0.61

 dishon

-0.61

orsche

-0.61

 delinquent

-0.60

 antis

-0.60

POSITIVE LOGITS

mere

1.20

wood

1.12

vale

0.98

seed

0.96

hurst

0.95

leigh

0.94

 Springs

0.91

stock

0.90

bridge

0.86

land

0.83

Activations Density 0.011%

references to specific locations or places, particularly those containing the word "wood"

references to locations or entities that include the word "wood."

No Comments

No Known Activations