INDEX

Explanations

references to legal actions and claims in lawsuits

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

iku

-0.76

ranean

-0.76

 Aires

-0.75

paren

-0.73

isSpecialOrderable

-0.73

dyl

-0.72

inav

-0.71

romeda

-0.69

yip

-0.69

uki

-0.69

POSITIVE LOGITS

 unlawfully

1.28

 violated

1.26

 negligent

1.20

 Defendants

1.20

 Defendant

1.17

 misrepresent

1.17

 Plaintiff

1.14

 neglig

1.13

 negligence

1.11

 defendants

1.10

Activations Density 0.221%

references to legal actions and claims in lawsuits

No Comments

No Known Activations