INDEX

Explanations

mentions of specific names and titles

oai_token-act-pair · gpt-3.5-turbo

mentions of specific individuals and their significance in a context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 redes

-0.67

 cryptocurrency

-0.63

DATA

-0.60

 watchdog

-0.59

 homeowner

-0.58

 HuffPost

-0.58

 uncover

-0.57

 prompts

-0.56

 autonom

-0.55

 urges

-0.55

POSITIVE LOGITS

 blah

1.02

 âĢ¦"

0.96

 ..."

0.95

['

0.95

 everybody

0.87

 anybody

0.87

["

0.87

 yeah

0.86

 somebody

0.83

 Godd

0.80

Activations Density 0.556%

mentions of specific names and titles

mentions of specific individuals and their significance in a context

No Comments

No Known Activations