INDEX

Explanations

verbs indicating past events or actions

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

ople

-0.69

ement

-0.64

 overpowered

-0.63

amn

-0.62

pack

-0.61

 overpower

-0.61

 stalls

-0.60

 fold

-0.59

tsy

-0.59

 arte

-0.59

POSITIVE LOGITS

 Lastly

1.14

Similarly

1.13

 Similarly

1.13

 Likewise

0.97

Lastly

0.93

 Finally

0.90

Likewise

0.87

 Meanwhile

0.85

 Another

0.82

Others

0.82

Activations Density 0.589%

verbs indicating past events or actions

No Comments

No Known Activations