INDEX

Explanations

phrases related to making predictions or deriving conclusions from data

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ndra

-0.75

VID

-0.73

natureconservancy

-0.70

 Carbuncle

-0.70

maid

-0.69

 bearer

-0.68

 Grail

-0.68

Interstitial

-0.67

 Teresa

-0.67

pee

-0.66

POSITIVE LOGITS

olated

1.44

olate

1.35

olation

1.28

 extrap

1.11

ol

0.89

ortal

0.88

ometric

0.88

ologies

0.87

olitics

0.85

oline

0.83

Activations Density 0.004%

phrases related to making predictions or deriving conclusions from data

No Comments

No Known Activations