INDEX

Explanations

phrases indicating a need for specific actions or items

oai_token-act-pair · gpt-3.5-turbo

statements indicating requirements or necessities

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

ð

-0.63

bridge

-0.58

Ń·

-0.56

Initialized

-0.56

resp

-0.55

 Seym

-0.54

riage

-0.53

atform

-0.53

threat

-0.51

berra

-0.51

POSITIVE LOGITS

to

1.06

lessly

1.02

0.86

to

0.85

 permission

0.77

 patience

0.71

TO

0.67

 assurances

0.66

 access

0.65

 authorization

0.64

Activations Density 0.065%

phrases indicating a need for specific actions or items

statements indicating requirements or necessities

No Comments

No Known Activations