INDEX

Explanations

phrases indicating a surprising or revealing discovery

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "it turns out."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

è¦ļéĨĴ

-1.05

accompan

-0.80

rongh

-0.70

riot

-0.68

 Expansion

-0.67

 Repeat

-0.64

aign

-0.59

-0.58

 Reserved

-0.57

riots

-0.57

POSITIVE LOGITS

out

1.08

entious

0.77

orned

0.66

 inward

0.65

out

0.65

outs

0.63

hift

0.62

enum

0.60

 forth

0.60

 doubtful

0.60

Activations Density 0.017%

phrases indicating a surprising or revealing discovery

instances of the phrase "it turns out."

No Comments

No Known Activations