INDEX

Explanations

terms related to misrepresentation and misinterpretation in various contexts

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

âĺħâĺħ

-0.59

 Mara

-0.58

bey

-0.58

 Dani

-0.57

 CARD

-0.57

 foremost

-0.56

 ashore

-0.56

STON

-0.56

ULTS

-0.56

ï¸

-0.55

POSITIVE LOGITS

ation

2.21

ations

1.86

eering

1.51

ational

1.49

ated

1.37

ed

1.32

ATIONS

1.27

ATION

1.26

ative

1.26

ationally

1.24

Activations Density 0.028%

terms related to misrepresentation and misinterpretation in various contexts

No Comments

No Known Activations