INDEX

Explanations

phrases related to legal or authoritative actions

oai_token-act-pair · gpt-3.5-turbo

references to legal proceedings and case outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

toggle

-0.79

 nerds

-0.77

 snag

-0.76

 perk

-0.74

 trendy

-0.73

 booming

-0.72

 quirky

-0.72

 pricey

-0.72

 locals

-0.72

 futuristic

-0.72

POSITIVE LOGITS

 Defendant

1.19

 Honour

1.14

 Plaintiff

1.12

 hereby

1.10

displayText

1.06

 appellant

1.03

 Defendants

0.99

Therefore

0.97

 Applicant

0.96

Chair

0.95

Activations Density 1.546%

phrases related to legal or authoritative actions

references to legal proceedings and case outcomes

No Comments

No Known Activations