INDEX

Explanations

proper nouns specifically related to the news organization "Bloomberg"

oai_token-act-pair · gpt-3.5-turbo

mentions of the Bloomberg news organization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

heid

-0.81

FUL

-0.81

mble

-0.78

avery

-0.77

netflix

-0.76

ulhu

-0.75

eem

-0.71

emale

-0.70

 Leia

-0.69

fully

-0.69

POSITIVE LOGITS

 Bloomberg

0.90

Business

0.82

Money

0.76

 News

0.75

 Sachs

0.75

ogl

0.74

Bloomberg

0.72

uggets

0.71

 exec

0.70

 Business

0.68

Activations Density 0.008%

proper nouns specifically related to the news organization "Bloomberg"

mentions of the Bloomberg news organization

No Comments

No Known Activations