INDEX

Explanations

words related to justification or acknowledgment

oai_token-act-pair · gpt-3.5-turbo

actions related to justifying, acknowledging, or integrating certain viewpoints or behaviors, particularly in complex social or political contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

talk

-0.47

behind

-0.44

NetMessage

-0.43

Ĥİ

-0.37

onsense

-0.36

si

-0.36

 Recon

-0.35

efer

-0.35

ternity

-0.35

POSITIVE LOGITS

allery

0.50

eering

0.48

issance

0.48

redients

0.46

tons

0.46

oaded

0.44

 conduc

0.43

ividual

0.42

ictionary

0.41

icate

0.40

Activations Density 0.740%

words related to justification or acknowledgment

actions related to justifying, acknowledging, or integrating certain viewpoints or behaviors, particularly in complex social or political contexts

No Comments

No Known Activations