INDEX

Explanations

phrases related to actions and decisions taken by organizations or authorities

oai_token-act-pair · gpt-3.5-turbo

entities and organizations involved in regulatory and investigative actions

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

cill

-0.73

 stones

-0.70

 grapes

-0.69

Interstitial

-0.66

 Georgian

-0.66

 amph

-0.66

die

-0.66

bou

-0.65

caps

-0.65

çĦ

-0.65

POSITIVE LOGITS

ngth

0.98

Department

0.88

 investigating

0.87

 Accountability

0.83

enhagen

0.82

 Department

0.82

 headquartered

0.81

 conducting

0.81

 guid

0.80

 oversight

0.80

Activations Density 0.539%

phrases related to actions and decisions taken by organizations or authorities

entities and organizations involved in regulatory and investigative actions

No Comments

No Known Activations