INDEX

Explanations

instances where actions are performed on or with the involvement of others

oai_token-act-pair · gpt-3.5-turbo

references to other individuals or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

aceous

-0.69

ISTER

-0.67

ister

-0.65

opy

-0.63

 Kitchen

-0.62

 Priest

-0.62

ories

-0.61

ropolis

-0.61

POSITIVE LOGITS

worldly

1.04

 behavi

1.01

 challeng

0.98

ĸļ

0.89

 describ

0.89

 redes

0.83

 harmed

0.80

 undermin

0.79

swer

0.79

 indo

0.77

Activations Density 0.023%

instances where actions are performed on or with the involvement of others

references to other individuals or entities

No Comments

No Known Activations