INDEX

Explanations

words related to morphological analysis, specifically looking for words ending in specific patterns like '-ing' and '-ation'

oai_token-act-pair · gpt-3.5-turbo

patterns related to words and their characteristics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

funding

-0.76

funded

-0.74

 sustainability

-0.73

 sandbox

-0.72

 starved

-0.72

 monitoring

-0.71

 starving

-0.71

 overhe

-0.70

vidia

-0.70

 freezer

-0.70

POSITIVE LOGITS

 pronunciation

1.61

 vowel

1.41

 suffix

1.40

tymology

1.38

 abbre

1.38

 noun

1.35

 slang

1.34

 pronouns

1.31

 notation

1.29

 prefix

1.27

Activations Density 0.833%

words related to morphological analysis, specifically looking for words ending in specific patterns like '-ing' and '-ation'

patterns related to words and their characteristics

No Comments

No Known Activations