INDEX

Explanations

informational terms or entities such as email addresses, URLs, or data about specific topics

oai_token-act-pair · gpt-3.5-turbo

instances of the word "info" or its variations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aws

-0.75

saf

-0.71

 reciprocal

-0.69

 agon

-0.69

 knees

-0.67

 friendships

-0.66

 frequent

-0.66

ŃĶ

-0.66

 tremend

-0.66

 forc

-0.66

POSITIVE LOGITS

info

1.08

Info

0.90

llor

0.90

etta

0.85

erences

0.84

anyahu

0.81

Information

0.79

etter

0.75

irmation

0.74

borough

0.73

Activations Density 0.011%

informational terms or entities such as email addresses, URLs, or data about specific topics

instances of the word "info" or its variations

No Comments

No Known Activations