INDEX

Explanations

mentions of governmental and political entities, positions, and actions

oai_token-act-pair · gpt-3.5-turbo

references to legislative processes and government bodies

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

maxwell

-0.85

 Courtney

-0.75

 Discovery

-0.74

Jones

-0.73

Sea

-0.71

 Keystone

-0.69

uckle

-0.68

Dave

-0.67

 pudding

-0.67

awk

-0.67

POSITIVE LOGITS

ÄŁ

1.03

 belonging

0.93

 Ankara

0.92

 deputies

0.91

 ministry

0.90

 decree

0.90

 parliament

0.89

 ministries

0.89

 municipality

0.89

 Syri

0.89

Activations Density 0.461%

mentions of governmental and political entities, positions, and actions

references to legislative processes and government bodies

No Comments

No Known Activations