INDEX

Explanations

instances of uncertainty or conditional phrases in discussions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Carn

-0.74

urous

-0.73

Runner

-0.69

 Carnival

-0.68

 Sailor

-0.68

 Hive

-0.67

ABE

-0.66

CCC

-0.65

esy

-0.65

 Journey

-0.64

POSITIVE LOGITS

but

0.81

butt

0.79

warts

0.75

okemon

0.75

BUT

0.74

bat

0.74

anwhile

0.73

 wart

0.73

wart

0.71

plets

0.71

Activations Density 0.110%

instances of uncertainty or conditional phrases in discussions

No Comments

No Known Activations