INDEX

Explanations

political and governmental references

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to political officials and their affiliations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

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

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_post

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

 Cherokee

-0.92

 Robertson

-0.84

 Chero

-0.79

 Griffin

-0.78

 Okin

-0.77

 Akin

-0.77

Yaz

-0.76

 Ramsay

-0.76

 Tanaka

-0.75

 Homer

-0.75

POSITIVE LOGITS

Calif

0.77

VT

0.73

democratic

0.72

California

0.71

0.70

achusetts

0.69

Israeli

0.68

pillar

0.68

0.67

Mich

0.67

Activations Density 0.053%

political and governmental references

references to political officials and their affiliations

No Comments

No Known Activations