INDEX

Explanations

legal and justice-related terms

oai_token-act-pair · gpt-3.5-turbo

references to legal issues and cinematic elements

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

cair

-0.89

umn

-0.81

NAS

-0.77

lish

-0.75

biz

-0.74

lik

-0.72

NYSE

-0.71

bent

-0.70

LIN

-0.68

ovation

-0.67

POSITIVE LOGITS

 whistleblowers

0.70

 protestors

0.67

 helpers

0.66

 protesters

0.65

weights

0.65

 holders

0.62

 supervisors

0.62

 gloves

0.61

 combatants

0.61

 Amendment

0.60

Activations Density 0.330%

legal and justice-related terms

references to legal issues and cinematic elements

No Comments

No Known Activations