INDEX

Explanations

keywords related to legal actions, specifically criminal charges being filed against individuals

oai_token-act-pair · gpt-3.5-turbo

phrases indicating legal charges or accusations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.90

aucuses

-0.85

soDeliveryDate

-0.83

ãĥ¼ãĥĨãĤ£

-0.80

ãĥ©ãĥ³

-0.79

 helicop

-0.77

 teasp

-0.76

ãĤĬ

-0.76

patch

-0.76

ãĤ¼ãĤ¦ãĤ¹

-0.74

POSITIVE LOGITS

him

0.94

 defendants

0.84

 alleged

0.81

 anyone

0.79

 whistleblowers

0.78

 anybody

0.74

 individuals

0.73

 wrongdoing

0.73

 innocent

0.73

 defendant

0.72

Activations Density 0.074%

keywords related to legal actions, specifically criminal charges being filed against individuals

phrases indicating legal charges or accusations

No Comments

No Known Activations