INDEX

Explanations

proper nouns identifying individuals in news reports

oai_token-act-pair · gpt-3.5-turbo

phrases that identify individuals or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

itiveness

-0.78

rones

-0.77

OTAL

-0.74

Duration

-0.70

ITH

-0.69

raq

-0.69

Length

-0.69

amon

-0.68

Capture

-0.66

imize

-0.66

POSITIVE LOGITS

 belonging

0.96

 follows

0.92

pired

0.80

 well

0.78

 favoring

0.78

pires

0.78

 opposed

0.76

 having

0.76

 conscientious

0.76

 being

0.70

Activations Density 0.103%

proper nouns identifying individuals in news reports

phrases that identify individuals or groups

No Comments

No Known Activations