INDEX

Explanations

past tense verbs that indicate completed actions

oai_token-act-pair · gpt-3.5-turbo

the word "been" indicating past actions or states

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

MO

-0.55

Plot

-0.55

 Slayer

-0.53

ule

-0.51

 Apart

-0.51

models

-0.51

ASH

-0.50

ooks

-0.50

NP

-0.50

ism

-0.49

POSITIVE LOGITS

 been

2.61

been

2.48

 Been

2.06

 gotten

1.84

 undergone

1.57

 gone

1.50

 begun

1.44

 arisen

1.39

 fallen

1.34

 become

1.32

Activations Density 0.148%

past tense verbs that indicate completed actions

the word "been" indicating past actions or states

No Comments

No Known Activations