INDEX

Explanations

terms related to unexpected or shocking events

oai_token-act-pair · gpt-3.5-turbo

phrases associated with surprising or shocking events and concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

verning

-0.81

 certific

-0.71

 GOODMAN

-0.68

 trustworthy

-0.67

 adequate

-0.64

veland

-0.64

 sufficient

-0.64

 proper

-0.62

 fullest

-0.61

versely

-0.61

POSITIVE LOGITS

ulence

0.77

fell

0.67

 conco

0.67

olor

0.67

 stunts

0.67

iasco

0.66

 Slov

0.64

 imaginable

0.63

 Vaugh

0.63

eries

0.61

Activations Density 0.473%

terms related to unexpected or shocking events

phrases associated with surprising or shocking events and concepts

No Comments

No Known Activations