INDEX

Explanations

terms related to prestigious organizations, rankings, and records

oai_token-act-pair · gpt-3.5-turbo

references to organizations and records related to health and cultural heritage

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

itars

-0.78

Lt

-0.78

amon

-0.75

 Hutch

-0.73

 McAuliffe

-0.71

 Boehner

-0.70

vous

-0.68

 Schumer

-0.67

SourceFile

-0.66

ulet

-0.64

POSITIVE LOGITS

 rankings

1.22

 criteria

1.09

 recommendations

1.07

 ratings

1.06

 rating

1.01

 classify

1.00

 recommendation

1.00

 Rankings

1.00

 guidelines

0.99

 Ratings

0.98

Activations Density 0.191%

terms related to prestigious organizations, rankings, and records

references to organizations and records related to health and cultural heritage

No Comments

No Known Activations