INDEX

Explanations

descriptions related to legal or political controversies

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

enhagen

-0.79

ãĤ´ãĥ³

-0.67

fman

-0.66

 Beware

-0.66

ãĥ¯

-0.66

 flush

-0.66

 reversible

-0.62

 scales

-0.62

 succeeding

-0.61

 multiplier

-0.61

POSITIVE LOGITS

anging

1.27

ospital

1.26

ISTORY

1.21

ometown

1.20

ottest

1.19

oused

1.17

umble

1.15

ollywood

1.15

annah

1.14

acking

1.13

Activations Density 0.284%

descriptions related to legal or political controversies

No Comments

No Known Activations