INDEX

Explanations

references to controversies and notable events in news and politics

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

 planner

-0.69

 halves

-0.66

 Sacrifice

-0.65

 Incarnation

-0.63

 Constitution

-0.63

 Assignment

-0.62

sac

-0.62

bis

-0.62

 terminated

-0.61

(?,

-0.61

POSITIVE LOGITS

 pundits

1.19

 critics

1.17

 commentators

1.13

 ridicule

1.07

Critics

1.04

 scorn

1.04

 detractors

1.02

 criticism

1.01

 skeptics

1.00

 acclaim

0.96

Activations Density 8.045%

references to controversies and notable events in news and politics

No Comments

No Known Activations