INDEX

Explanations

proper nouns or headlines in a news article

oai_token-act-pair · gpt-3.5-turbo

multiple occurrences of the word "Story" used in a series of headlines or summaries

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

ergic

-0.71

 consecut

-0.63

ullivan

-0.61

 indiscrim

-0.59

 respectively

-0.58

 simultaneously

-0.56

iate

-0.55

oner

-0.55

ktop

-0.54

ãĥĺãĥ©

-0.52

POSITIVE LOGITS

telling

1.22

 highlights

1.16

 Highlights

1.08

 Transcript

0.90

 Continued

0.82

line

0.82

stal

0.81

lines

0.81

anooga

0.78

summary

0.77

Activations Density 0.033%

proper nouns or headlines in a news article

multiple occurrences of the word "Story" used in a series of headlines or summaries

No Comments

No Known Activations