INDEX

Explanations

mentions of news coverage or stories

oai_token-act-pair · gpt-3.5-turbo

elements related to media coverage and news stories

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

 palate

-0.74

ertodd

-0.74

*/(

-0.72

ranch

-0.70

inence

-0.68

 joystick

-0.67

ojure

-0.67

orest

-0.66

ometry

-0.65

gettable

-0.65

POSITIVE LOGITS

reporting

0.86

MSM

0.81

 Murdoch

0.80

 reporting

0.80

 articles

0.77

essler

0.75

 dissemin

0.74

 headlines

0.74

reports

0.74

 headlined

0.73

Activations Density 0.154%

mentions of news coverage or stories

elements related to media coverage and news stories

No Comments

No Known Activations