INDEX

Explanations

directional words and phrases indicating movement or position

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

rams

-0.67

ivities

-0.67

ories

-0.63

usc

-0.62

atively

-0.61

xious

-0.60

sels

-0.60

uren

-0.58

iny

-0.58

matically

-0.57

POSITIVE LOGITS

cliffe

0.70

ruary

0.67

stage

0.62

 Vulcan

0.62

ategory

0.60

stairs

0.60

othal

0.60

hovah

0.60

flix

0.59

WARD

0.59

Activations Density 0.011%

directional words and phrases indicating movement or position

No Comments

No Known Activations