INDEX

Explanations

phrases related to conflicts or arguments

oai_token-act-pair · gpt-3.5-turbo

instances of conflict or fighting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

eele

-0.77

engineering

-0.74

velt

-0.71

 stocking

-0.71

akeru

-0.70

imate

-0.65

icrobial

-0.64

prints

-0.64

ographic

-0.63

ocl

-0.62

POSITIVE LOGITS

 ensued

1.10

 between

0.99

 raging

0.88

 brawl

0.88

 erupted

0.86

 raged

0.84

SHIP

0.82

 disagreement

0.82

 involving

0.82

 banter

0.81

Activations Density 0.209%

phrases related to conflicts or arguments

instances of conflict or fighting

No Comments

No Known Activations