INDEX

Explanations

highly belligerent and confrontational language

oai_token-act-pair · gpt-3.5-turbo

terms related to conflicts or warfare

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.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

 Tammy

-0.64

 chops

-0.64

 trophies

-0.61

 bills

-0.61

 Lake

-0.59

 Mozilla

-0.59

Lake

-0.59

 hoped

-0.58

 hugs

-0.57

sm

-0.57

POSITIVE LOGITS

erent

4.91

eren

1.24

erential

1.23

arent

1.11

erence

1.07

iliar

1.05

arant

1.04

ividual

1.02

minist

1.00

erest

0.98

Activations Density 0.013%

highly belligerent and confrontational language

terms related to conflicts or warfare

No Comments

No Known Activations