INDEX

Explanations

mentions related to the TV network NBC

oai_token-act-pair · gpt-3.5-turbo

mentions of the NBC network

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

aunder

-0.73

andr

-0.71

erva

-0.70

 McCartney

-0.69

cific

-0.69

piration

-0.68

rosse

-0.67

Riy

-0.66

arians

-0.64

ãĤ¢ãĥ«

-0.64

POSITIVE LOGITS

Universal

1.25

NBC

0.85

SN

0.84

 affiliate

0.84

WS

0.78

 News

0.74

NBC

0.74

 Networks

0.72

 Universal

0.72

 affiliates

0.71

Activations Density 0.009%

mentions related to the TV network NBC

mentions of the NBC network

No Comments

No Known Activations