INDEX

Explanations

terms related to military personnel and units, particularly focusing on infantry and cavalry

oai_token-act-pair · gpt-3.5-turbo

references to military forces and terms

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

à

-0.84

é¾įå¥ĳå£«

-0.74

å¥³

-0.69

 Prosper

-0.68

________________________

-0.68

åħ

-0.64

________________

-0.63

 Neighbor

-0.62

uras

-0.60

âĸ¬

-0.60

POSITIVE LOGITS

 infantry

1.03

 colonel

0.94

layer

0.88

ensibly

0.83

 battalion

0.82

 barracks

0.82

 barr

0.80

 Infantry

0.80

rats

0.77

 cavalry

0.76

Activations Density 0.008%

terms related to military personnel and units, particularly focusing on infantry and cavalry

references to military forces and terms

No Comments

No Known Activations