INDEX

Explanations

keywords related to legal or political actions and decisions

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

aughs

-0.69

 sucks

-0.61

Simple

-0.58

iversary

-0.57

asus

-0.56

 Circus

-0.56

rics

-0.55

IFE

-0.54

bye

-0.54

VID

-0.54

POSITIVE LOGITS

ourcing

1.04

 ranging

0.98

hooting

0.97

 pertaining

0.97

 relating

0.96

afety

0.94

hops

0.93

pread

0.92

cape

0.89

ynthesis

0.88

Activations Density 0.312%

keywords related to legal or political actions and decisions

No Comments

No Known Activations