INDEX

Explanations

references to military personnel or staff

oai_token-act-pair · gpt-3.5-turbo

references to personnel in a military or organizational context

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

bows

-0.89

bow

-0.86

forth

-0.82

vid

-0.78

nings

-0.77

TERN

-0.75

bolt

-0.73

night

-0.70

ibia

-0.68

ray

-0.67

POSITIVE LOGITS

 stationed

0.94

 personnel

0.90

ilitary

0.87

 secretaries

0.86

atics

0.85

 carriers

0.84

geant

0.82

ocus

0.78

 corps

0.78

atcher

0.75

Activations Density 0.012%

references to military personnel or staff

references to personnel in a military or organizational context

No Comments

No Known Activations