INDEX

Explanations

negations and negated actions

oai_token-act-pair · gpt-3.5-turbo

negative constructions and phrases indicating absence or non-existence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Advice

-0.68

 ANGEL

-0.66

DIT

-0.66

 Statements

-0.65

éĥ

-0.65

CAN

-0.65

çĶŁ

-0.64

 interim

-0.64

 Moves

-0.64

 Lessons

-0.63

POSITIVE LOGITS

 otherwise

1.05

icable

1.01

 necessarily

1.01

fit

0.83

fit

0.82

yet

0.82

 ordinarily

0.82

necess

0.81

icably

0.81

 adequately

0.80

Activations Density 0.164%

negations and negated actions

negative constructions and phrases indicating absence or non-existence

No Comments

No Known Activations