INDEX

Explanations

references to specific decades, particularly the 1960s, 1970s, 1980s, and 1990s

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

venge

-0.65

ngth

-0.61

terday

-0.61

aukee

-0.61

 masc

-0.60

Tickets

-0.59

cam

-0.57

semble

-0.56

cham

-0.56

netflix

-0.56

POSITIVE LOGITS

1.19

-'

0.85

sie

0.85

eties

0.81

ixties

0.81

ies

0.80

ties

0.79

enthal

0.78

era

0.78

sburg

0.76

Activations Density 0.029%

references to specific decades, particularly the 1960s, 1970s, 1980s, and 1990s

No Comments

No Known Activations