INDEX

Explanations

references to military personnel, particularly soldiers

oai_token-act-pair · gpt-3.5-turbo

mentions of soldiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

TERN

-0.85

Merit

-0.69

Pwr

-0.69

York

-0.68

given

-0.67

Ħ¢

-0.66

 Neighbor

-0.65

omial

-0.65

lihood

-0.64

 Venus

-0.64

POSITIVE LOGITS

 fatig

0.97

 stationed

0.97

 soldiers

0.96

alions

0.85

 uniforms

0.84

 garrison

0.81

 acad

0.81

rats

0.80

 academy

0.78

 barracks

0.77

Activations Density 0.012%

references to military personnel, particularly soldiers

mentions of soldiers

No Comments

No Known Activations