INDEX

Explanations

specific references to political administrations, in particular the Obama and Trump administrations

oai_token-act-pair · gpt-3.5-turbo

references to political administrations, particularly Obama and Trump

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

aterasu

-0.83

NetMessage

-0.71

agree

-0.67

...]

-0.63

verbs

-0.63

 Takeru

-0.63

ocrates

-0.62

duty

-0.61

reddits

-0.61

ubb

-0.59

POSITIVE LOGITS

 administration

1.45

 Administration

1.29

 presidency

1.23

era

1.06

 campaign

1.05

 regime

1.03

 agenda

0.99

 playbook

0.99

 administrations

0.97

 dynasty

0.92

Activations Density 0.072%

specific references to political administrations, in particular the Obama and Trump administrations

references to political administrations, particularly Obama and Trump

No Comments

No Known Activations