INDEX

Explanations

terms related to denial or refutation of specific claims or allegations

oai_token-act-pair · gpt-3.5-turbo

phrases related to denial and disputes over evidence or claims

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

anders

-0.83

aunts

-0.81

ayers

-0.73

dt

-0.72

wered

-0.68

amph

-0.68

requently

-0.68

 partName

-0.67

Zup

-0.67

 Parables

-0.65

POSITIVE LOGITS

 connection

1.43

 wrongdoing

1.41

 nexus

1.38

 collusion

1.36

 motive

1.28

 link

1.28

 complicity

1.24

 conspiracy

1.21

 relationship

1.18

 violation

1.18

Activations Density 0.243%

terms related to denial or refutation of specific claims or allegations

phrases related to denial and disputes over evidence or claims

No Comments

No Known Activations