INDEX

Explanations

words related to actions or events

oai_token-act-pair · gpt-3.5-turbo

actions and their consequences

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

é¾į

-0.64

Eastern

-0.60

emen

-0.59

eu

-0.58

otonin

-0.58

Enlarge

-0.58

 scares

-0.57

iates

-0.57

ravity

-0.56

ymph

-0.56

POSITIVE LOGITS

ardless

0.74

ifully

0.70

akespeare

0.65

ãĤ¤ãĥĪ

0.65

igs

0.65

ERY

0.64

umblr

0.64

=#

0.63

 captcha

0.63

ãĥķãĤ¡

0.62

Activations Density 0.351%

words related to actions or events

actions and their consequences

No Known Activations