INDEX

Explanations

references to newsletters belonging to a specific entity

oai_token-act-pair · gpt-3.5-turbo

references to subscriptions or newsletters

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

nesota

-0.90

okemon

-0.77

 Izan

-0.77

minster

-0.73

wic

-0.70

stood

-0.69

oldemort

-0.68

teness

-0.68

',"

-0.67

-0.65

POSITIVE LOGITS

selves

0.93

own

0.93

 newest

0.87

 exclusive

0.82

 motto

0.80

 latest

0.80

 handy

0.79

 estimation

0.76

 inaugural

0.76

 fearless

0.75

Activations Density 0.094%

references to newsletters belonging to a specific entity

references to subscriptions or newsletters

No Comments

No Known Activations