INDEX

Explanations

phrases related to supporting or refuting research findings

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate support or validation of certain viewpoints or theories

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

sembly

-0.85

 actionGroup

-0.85

 toget

-0.83

abouts

-0.79

Cod

-0.75

phabet

-0.74

yip

-0.70

Events

-0.69

atra

-0.64

meet

-0.64

POSITIVE LOGITS

 hypothesis

1.57

 claim

1.56

 assertion

1.54

 notion

1.54

 contention

1.51

 premise

1.44

 theory

1.38

 thesis

1.38

 assumption

1.34

 proposition

1.34

Activations Density 0.309%

phrases related to supporting or refuting research findings

phrases that indicate support or validation of certain viewpoints or theories

No Comments

No Known Activations