INDEX

Explanations

legal terms related to rules, violations, regulations, and crimes

oai_token-act-pair · gpt-3.5-turbo

terms and phrases that define or classify actions, particularly in legal or regulatory contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ãĤ¼ãĤ¦ãĤ¹

-0.76

Ö¼

-0.74

 Tune

-0.72

onen

-0.69

 rhythms

-0.68

eters

-0.67

uckland

-0.67

angles

-0.66

enium

-0.66

many

-0.66

POSITIVE LOGITS

 obstruction

1.05

 genocide

0.97

 proof

0.97

 evidence

0.96

 treason

0.96

 endorsement

0.94

 conclusive

0.94

 disqual

0.92

 discrimination

0.91

 infringement

0.90

Activations Density 0.202%

legal terms related to rules, violations, regulations, and crimes

terms and phrases that define or classify actions, particularly in legal or regulatory contexts

No Comments

No Known Activations