INDEX

Explanations

references to the television network NBC

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "NBC."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

erva

-0.79

wikipedia

-0.78

antly

-0.75

cific

-0.73

rosse

-0.70

llan

-0.70

aunder

-0.69

uming

-0.69

arians

-0.69

piration

-0.69

POSITIVE LOGITS

Universal

1.37

SN

1.03

 News

0.98

News

0.93

 Sports

0.91

 SPORTS

0.86

conn

0.82

NBC

0.81

Sports

0.81

 affiliate

0.81

Activations Density 0.014%

references to the television network NBC

occurrences of the word "NBC."

No Comments

No Known Activations