INDEX

Explanations

statements referring to people or readers

oai_token-act-pair · gpt-3.5-turbo

references to individuals or groups of people in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

xon

-0.67

cknow

-0.65

 Correct

-0.65

saf

-0.61

ust

-0.60

rontal

-0.60

illus

-0.60

uphem

-0.59

amaru

-0.59

bin

-0.58

POSITIVE LOGITS

 aspiring

1.08

 alike

0.99

 wanting

0.96

 sake

0.96

erning

0.88

ãĤ®

0.86

 wishing

0.82

 needing

0.82

 purposes

0.78

 interested

0.77

Activations Density 0.223%

statements referring to people or readers

references to individuals or groups of people in various contexts

No Comments

No Known Activations