INDEX

Explanations

phrases and terms related to moral and legal judgments

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

acan

-0.68

iping

-0.67

nants

-0.67

rongh

-0.66

arta

-0.65

jong

-0.64

phabet

-0.62

-0.61

afia

-0.60

axy

-0.59

POSITIVE LOGITS

 raining

0.87

ceivable

0.75

 impossible

0.72

 folly

0.71

 coincidence

0.69

ifiable

0.65

to

0.64

 advisable

0.63

 conceivable

0.62

EC

0.62

Activations Density 0.269%

phrases and terms related to moral and legal judgments

No Comments

No Known Activations