INDEX

Explanations

instances of the word "int" and its variations, indicating a focus on intensity or measurement-related terms

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

è¦ļéĨĴ

-0.79

DT

-0.68

days

-0.67

phy

-0.66

phia

-0.65

brates

-0.65

pload

-0.65

 millenn

-0.64

BLE

-0.64

POSITIVE LOGITS

ypes

1.10

ellect

0.91

eki

0.88

aro

0.85

zman

0.83

opia

0.83

ured

0.82

ered

0.82

illation

0.81

essential

0.81

Activations Density 0.005%

instances of the word "int" and its variations, indicating a focus on intensity or measurement-related terms

No Comments

No Known Activations