INDEX

Explanations

mentions of news sources and reported information

oai_token-act-pair · gpt-3.5-turbo

sentences that report information or statements from various sources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 utter

-0.69

 ourselves

-0.66

 landscape

-0.65

à¼

-0.64

¬¼

-0.63

 elephant

-0.62

 pudding

-0.60

 vict

-0.60

 transistor

-0.60

 deserved

-0.60

POSITIVE LOGITS

 Notably

1.03

 Additionally

1.00

 Investigators

0.95

 However

0.93

 Specifically

0.90

 Meanwhile

0.89

 Another

0.89

 Similarly

0.86

 Presumably

0.86

 Along

0.86

Activations Density 0.279%

mentions of news sources and reported information

sentences that report information or statements from various sources

No Comments

No Known Activations