INDEX

Explanations

legal and investigative terms related to law enforcement and processes

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 typing

-0.66

 representations

-0.63

essor

-0.62

 pockets

-0.61

 clutter

-0.61

lder

-0.61

 candy

-0.61

stadt

-0.61

oak

-0.60

veland

-0.59

POSITIVE LOGITS

 underway

1.38

 undertaken

1.14

 occurring

1.13

 conducted

1.11

 unfolding

1.10

 involving

1.07

 happening

1.04

 unfold

0.99

 postponed

0.98

 waged

0.98

Activations Density 4.337%

legal and investigative terms related to law enforcement and processes

No Comments

No Known Activations