INDEX

Explanations

proper nouns or named entities mentioned in written text

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "to" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 pitted

-0.69

 deductions

-0.66

ynski

-0.65

 concentrated

-0.64

 botched

-0.63

 tracked

-0.62

 risks

-0.61

 fung

-0.61

 disadvantages

-0.60

 disproportionate

-0.60

POSITIVE LOGITS

wered

1.07

ilet

0.99

pless

0.97

adies

0.89

 accompany

0.89

ggles

0.88

 satisfy

0.88

psy

0.88

 appease

0.87

othy

0.85

Activations Density 0.315%

proper nouns or named entities mentioned in written text

occurrences of the word "to" in various contexts

No Comments

No Known Activations