INDEX

Explanations

words with the prefix "pre-" indicating something ahead of or before something else

oai_token-act-pair · gpt-3.5-turbo

prefixes and suffixes commonly used in word formation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ´ãĥ³

-0.61

 Skydragon

-0.60

 Dragonbound

-0.57

 Spears

-0.57

 omission

-0.55

 Watch

-0.54

 Targ

-0.54

lag

-0.53

 disav

-0.52

 boycot

-0.52

POSITIVE LOGITS

ulum

0.88

aturally

0.88

incial

0.87

essional

0.85

opausal

0.78

ilial

0.78

ieved

0.77

ophile

0.75

ule

0.75

essing

0.74

Activations Density 0.202%

words with the prefix "pre-" indicating something ahead of or before something else

prefixes and suffixes commonly used in word formation

No Comments

No Known Activations