INDEX

Explanations

instances where the text transitions and introduces new information

oai_token-act-pair · gpt-3.5-turbo

phrases starting with "This" that lead into varied contexts or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

adle

-0.72

pots

-0.66

icons

-0.65

ãĤ¹ãĥĪ

-0.63

masters

-0.63

ques

-0.62

aws

-0.60

ãĥĺ

-0.60

ivas

-0.60

aments

-0.60

POSITIVE LOGITS

 arrang

0.76

 week

0.75

information

0.74

 Month

0.73

 month

0.73

 trope

0.72

 discrepancy

0.70

 Week

0.69

 stoked

0.69

 newfound

0.69

Activations Density 0.088%

instances where the text transitions and introduces new information

phrases starting with "This" that lead into varied contexts or statements

No Comments

No Known Activations