INDEX

Explanations

expressions of disapproval or critique towards certain actions or policies

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

 nonetheless

-0.96

etheless

-0.78

 nevertheless

-0.68

ccording

-0.64

Mub

-0.62

 Lastly

-0.61

 Finally

-0.61

 instead

-0.60

ague

-0.60

 Minotaur

-0.58

POSITIVE LOGITS

 aesthetics

0.63

ifiable

0.61

 aest

0.59

erves

0.58

omical

0.58

 paycheck

0.57

 pian

0.57

 aesthetic

0.57

})

0.56

omic

0.55

Activations Density 0.113%

expressions of disapproval or critique towards certain actions or policies

No Comments

No Known Activations