INDEX

Explanations

adverbs indicating hidden or secret actions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "secretly" reflecting covert actions or hidden behaviors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 GOODMAN

-0.83

arcity

-0.83

nesota

-0.79

amins

-0.78

ordan

-0.75

Pwr

-0.75

 è£ıç

-0.74

alon

-0.74

amaz

-0.73

inez

-0.71

POSITIVE LOGITS

 tucked

0.76

 annex

0.76

 stash

0.75

 whispering

0.74

 sneak

0.73

 incest

0.71

 peek

0.70

 plotting

0.70

bom

0.69

 whispers

0.69

Activations Density 0.064%

adverbs indicating hidden or secret actions

instances of the word "secretly" reflecting covert actions or hidden behaviors

No Comments

No Known Activations