INDEX

Explanations

the verb "is" and its forms as indicators of state or condition

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

istries

-0.87

enaries

-0.71

acements

-0.68

 Ends

-0.65

blocks

-0.64

acement

-0.63

interstitial

-0.62

 Blocks

-0.61

 IMAGES

-0.61

eering

-0.60

POSITIVE LOGITS

 evidenced

1.19

 customary

1.08

 evident

1.04

 usual

0.95

 common

0.89

 doubtless

0.86

 abundantly

0.84

 rumored

0.82

 implied

0.81

 demonstrated

0.79

Activations Density 0.008%

the verb "is" and its forms as indicators of state or condition

No Comments

No Known Activations