INDEX

Explanations

specific years and dates mentioned in the text

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

Els

-0.67

Tam

-0.61

 hots

-0.60

 locations

-0.60

 rooft

-0.60

pread

-0.59

 lifes

-0.56

etimes

-0.56

paio

-0.55

 deserts

-0.54

POSITIVE LOGITS

0.79

ruary

0.79

 semester

0.77

/,

0.72

 season

0.70

 onwards

0.69

\.

0.69

-'

0.64

 timeframe

0.63

iatus

0.62

Activations Density 0.068%

specific years and dates mentioned in the text

No Comments

No Known Activations