INDEX

Explanations

the word "Yet" 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

tained

-0.80

ãĥİ

-0.80

ãĤ¼ãĤ¦ãĤ¹

-0.75

ding

-0.70

omial

-0.70

tein

-0.70

hal

-0.69

ãĤ¡

-0.69

ga

-0.68

packs

-0.67

POSITIVE LOGITS

 somehow

1.03

 despite

0.90

 strangely

0.90

tons

0.85

 interestingly

0.78

 alas

0.77

 nonetheless

0.77

 paradox

0.75

 persisted

0.75

withstanding

0.74

Activations Density 0.011%

the word "Yet" in various contexts

No Comments

No Known Activations