INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names, particularly related to individuals and managers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

--------------------------------------------------------

-0.66

 Springs

-0.66

ollar

-0.65

edo

-0.65

CBI

-0.64

ragon

-0.63

 Wolves

-0.62

 liner

-0.62

plan

-0.61

LD

-0.60

POSITIVE LOGITS

iversal

0.95

atform

0.87

hem

0.85

ishi

0.83

imil

0.75

acebook

0.75

inas

0.74

issan

0.72

anson

0.70

sand

0.68

Activations Density 0.032%

names of individuals

mentions of specific names, particularly related to individuals and managers

No Comments

No Known Activations