INDEX

Explanations

instances of the word "excess"

oai_token-act-pair · gpt-3.5-turbo

references to the concept of excess or surplus

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

udder

-0.81

ramid

-0.78

hran

-0.73

herer

-0.73

osuke

-0.73

chens

-0.69

mberg

-0.68

HCR

-0.67

wark

-0.67

 Brist

-0.67

POSITIVE LOGITS

 excess

0.99

 baggage

0.86

 overfl

0.80

eatures

0.78

 workload

0.72

 amounts

0.72

 accumulation

0.71

 wast

0.70

atile

0.69

 negativity

0.69

Activations Density 0.005%

instances of the word "excess"

references to the concept of excess or surplus

No Comments

No Known Activations