INDEX

Explanations

proper nouns such as names of companies or organizations

oai_token-act-pair · gpt-3.5-turbo

acronyms, abbreviations, or initialisms related to various organizations or topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

enegger

-0.93

 disputed

-0.83

 bottleneck

-0.83

 stroke

-0.77

 dividing

-0.75

 recess

-0.75

 confined

-0.75

 charge

-0.74

 length

-0.73

 substance

-0.71

POSITIVE LOGITS

Fest

1.03

GN

0.99

Forge

0.97

FM

0.96

Week

0.95

GF

0.95

WN

0.93

BW

0.92

TV

0.92

OTA

0.91

Activations Density 0.396%

proper nouns such as names of companies or organizations

acronyms, abbreviations, or initialisms related to various organizations or topics

No Comments

No Known Activations