INDEX

Explanations

words related to changes in numerical values or quantities

oai_token-act-pair · gpt-3.5-turbo

quantifiable changes or trends in data, specifically focusing on the frequency of occurrences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

leon

-0.77

çīĪ

-0.75

woods

-0.73

ĸļ

-0.73

æ³

-0.72

blind

-0.71

fuck

-0.71

Planet

-0.71

bol

-0.70

raid

-0.70

POSITIVE LOGITS

 popularity

1.13

 productivity

1.02

 revenues

0.98

 visitation

0.97

 profitability

0.96

 sales

0.96

 incidence

0.95

 activity

0.94

 attendance

0.93

 competitiveness

0.92

Activations Density 0.095%

words related to changes in numerical values or quantities

quantifiable changes or trends in data, specifically focusing on the frequency of occurrences

No Comments

No Known Activations