INDEX

Explanations

years or time periods in phrases

oai_token-act-pair · gpt-3.5-turbo

references to specific years and financial figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 desks

-0.77

 scanner

-0.65

 contr

-0.65

Free

-0.62

 corners

-0.61

Sov

-0.61

 TEXT

-0.60

Dialogue

-0.60

 antit

-0.60

igraph

-0.59

POSITIVE LOGITS

 onwards

0.98

 onward

0.75

etheus

0.73

govtrack

0.71

netflix

0.68

ndra

0.67

teness

0.64

 Acceler

0.64

 Cance

0.63

hua

0.63

Activations Density 0.359%

years or time periods in phrases

references to specific years and financial figures

No Comments

No Known Activations