INDEX

Explanations

phrases describing controversy or conflict

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

inction

-0.77

verage

-0.76

clips

-0.75

redits

-0.71

estyles

-0.69

apest

-0.69

wcsstore

-0.69

 largeDownload

-0.69

uality

-0.68

served

-0.67

POSITIVE LOGITS

 superiors

1.26

 stakeholders

1.18

 policymakers

1.11

 representatives

1.11

 legislators

1.08

 passers

1.07

 locals

1.06

 regulators

1.06

 employers

1.06

 landowners

1.06

Activations Density 2.957%

phrases describing controversy or conflict

No Comments

No Known Activations