INDEX

Explanations

words related to historical events and places

oai_token-act-pair · gpt-3.5-turbo

references to institutions, events, and formal structures in society

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

ĻĤ

-0.81

terday

-0.75

ĪĴ

-0.70

 traumatic

-0.61

iannopoulos

-0.61

serving

-0.58

OUP

-0.57

 nonex

-0.57

 karma

-0.56

etting

-0.56

POSITIVE LOGITS

Of

1.01

 Council

0.93

 Gazette

0.92

 Builder

0.92

 Accountability

0.90

 Chamber

0.90

 Room

0.89

 Wars

0.89

 Committee

0.88

 Fathers

0.86

Activations Density 0.200%

words related to historical events and places

references to institutions, events, and formal structures in society

No Comments

No Known Activations