INDEX

Explanations

terms related to politics and policy, with a focus on social justice, mental health, and government 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_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

ãĤ¨ãĥ«

-0.96

éŃĶ

-0.83

SN

-0.81

ãĥĥ

-0.81

stown

-0.79

ãĥĥãĥī

-0.77

GMT

-0.75

ãĥ¥

-0.74

odox

-0.74

gro

-0.74

POSITIVE LOGITS

 superpower

0.83

eers

0.76

 equivalents

0.72

 nonprofits

0.69

 pamph

0.68

 dystop

0.68

 issues

0.68

 optimization

0.68

 guru

0.67

 satire

0.67

Activations Density 2.078%

terms related to politics and policy, with a focus on social justice, mental health, and government actions

No Comments

No Known Activations