INDEX

Explanations

phrases indicating sharing or reading stories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

DERR

-0.64

calling

-0.61

gow

-0.61

stood

-0.61

rats

-0.61

 reeling

-0.60

taking

-0.60

gem

-0.59

pher

-0.57

 domest

-0.57

POSITIVE LOGITS

 partName

0.67

 PsyNet

0.65

imon

0.64

iframe

0.63

 Tribune

0.63

 Expand

0.62

 LINK

0.61

 annotation

0.60

ascar

0.59

 story

0.58

Activations Density 0.019%

phrases indicating sharing or reading stories

No Comments

No Known Activations